Finite Element Modeling and Simulation with ANSYS Workbench-CRC Press (2014)
advertisement
Xiaolin Chen • Yijun Liu
Xiaolin Chen • Yijun Liu
Boca Raton London New York
CRC Press is an imprint of the
Taylor & Francis Group, an informa business
ANSYS, ANSYS Workbench, Ansoft, AUTODYN, CFX, EKM, Engineering Knowledge Manager, FLUENT, HFSS, and any and all ANSYS, Inc. brand, product, service and feature names, logos and slogans
are trademarks or registered trademarks of ANSYS, Inc. or its subsidiaries located in the United States
or other countries. ICEM CFD is a trademark used by ANSYS, Inc. under license. CFX is a trademark
of Sony Corporation in Japan. All other brand, product, service, and feature names or trademarks are
the property of their respective companies.
CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2015 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Version Date: 20140707
International Standard Book Number-13: 978-1-4398-7385-4 (eBook - PDF)
This book contains information obtained from authentic and highly regarded sources. Reasonable
efforts have been made to publish reliable data and information, but the author and publisher cannot
assume responsibility for the validity of all materials or the consequences of their use. The authors and
publishers have attempted to trace the copyright holders of all material reproduced in this publication
and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any
future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222
Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are
used only for identification and explanation without intent to infringe.
Visit the Taylor & Francis Web site at
http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
Contents
Preface...............................................................................................................................................xi
Authors.......................................................................................................................................... xiii
1. Introduction..............................................................................................................................1
1.1 Some Basic Concepts.....................................................................................................1
1.1.1 Why FEA?..........................................................................................................1
1.1.2 Finite Element Applications in Engineering.................................................1
1.1.3 FEA with ANSYS Workbench...........................................................................3
1.1.4 A Brief History of FEA..................................................................................... 3
1.1.5 A General Procedure for FEA......................................................................... 4
1.2 An Example in FEA: Spring System............................................................................4
1.2.1 One Spring Element.........................................................................................5
1.2.2 A Spring System................................................................................................ 6
1.2.2.1 Assembly of Element Equations: Direct Approach......................6
1.2.2.2 Assembly of Element Equations: Energy Approach.................... 8
1.2.3 Boundary and Load Conditions..................................................................... 9
1.2.4 Solution Verification....................................................................................... 10
1.2.5 Example Problems.......................................................................................... 10
1.3 Overview of ANSYS Workbench................................................................................. 13
1.3.1 The User Interface........................................................................................... 13
1.3.2 The Toolbox...................................................................................................... 14
1.3.3 The Project Schematic.................................................................................... 14
1.3.4 Working with Cells......................................................................................... 16
1.3.5 The Menu Bar.................................................................................................. 16
1.4 Summary....................................................................................................................... 17
Problems................................................................................................................................... 18
2. Bars and Trusses.................................................................................................................... 21
2.1 Introduction.................................................................................................................. 21
2.2 Review of the 1-D Elasticity Theory.......................................................................... 21
2.3 Modeling of Trusses....................................................................................................22
2.4 Formulation of the Bar Element................................................................................. 23
2.4.1 Stiffness Matrix: Direct Method................................................................... 23
2.4.2 Stiffness Matrix: Energy Approach.............................................................. 25
2.4.3 Treatment of Distributed Load..................................................................... 27
2.4.4 Bar Element in 2-D and 3-D........................................................................... 28
2.4.4.1 2-D Case............................................................................................ 28
2.4.4.2 3-D Case............................................................................................ 31
2.4.5 Element Stress................................................................................................. 31
2.5 Examples with Bar Elements...................................................................................... 32
2.6 Case Study with ANSYS Workbench.......................................................................... 40
2.7 Summary....................................................................................................................... 52
2.8 Review of Learning Objectives.................................................................................. 52
Problems................................................................................................................................... 52
v
vi
Contents
3. Beams and Frames................................................................................................................. 57
3.1 Introduction.................................................................................................................. 57
3.2 Review of the Beam Theory....................................................................................... 57
3.2.1 Euler–Bernoulli Beam and Timoshenko Beam.......................................... 57
3.2.2 Stress, Strain, Deflection, and Their Relations........................................... 59
3.3 Modeling of Beams and Frames................................................................................ 60
3.3.1 Cross Sections and Strong/Weak Axis........................................................ 60
3.3.2 Support Conditions........................................................................................ 61
3.3.3 Conversion of a Physical Model into a Line Model................................... 62
3.4 Formulation of the Beam Element............................................................................. 62
3.4.1 Element Stiffness Equation: The Direct Approach....................................63
3.4.2 Element Stiffness Equation: The Energy Approach..................................64
3.4.3 Treatment of Distributed Loads.................................................................... 66
3.4.4 Stiffness Matrix for a General Beam Element............................................. 67
3.5 Examples with Beam Elements.................................................................................. 68
3.6 Case Study with ANSYS Workbench..........................................................................77
3.7 Summary....................................................................................................................... 96
3.8 Review of Learning Objectives.................................................................................. 96
Problems................................................................................................................................... 96
4. Two-Dimensional Elasticity.............................................................................................. 101
4.1 Introduction................................................................................................................ 101
4.2 Review of 2-D Elasticity Theory.............................................................................. 101
4.2.1 Plane Stress.................................................................................................... 101
4.2.2 Plane Strain.................................................................................................... 102
4.2.3 Stress–Strain (Constitutive) Equations...................................................... 103
4.2.4 Strain and Displacement Relations............................................................ 104
4.2.5 Equilibrium Equations................................................................................. 105
4.2.6 Boundary Conditions................................................................................... 105
4.2.7 Exact Elasticity Solution............................................................................... 105
4.3 Modeling of 2-D Elasticity Problems....................................................................... 106
4.4 Formulation of the Plane Stress/Strain Element................................................... 107
4.4.1 A General Formula for the Stiffness Matrix............................................. 108
4.4.2 Constant Strain Triangle (CST or T3)......................................................... 108
4.4.3 Quadratic Triangular Element (LST or T6)............................................... 113
4.4.4 Linear Quadrilateral Element (Q4)............................................................. 114
4.4.5 Quadratic Quadrilateral Element (Q8)....................................................... 115
4.4.6 Transformation of Loads.............................................................................. 116
4.4.7 Stress Calculation......................................................................................... 118
4.4.7.1 The von Mises Stress.................................................................... 118
4.4.7.2 Averaged Stresses.......................................................................... 119
4.4.8 General Comments on the 2-D Elements.................................................. 120
4.5 Case Study with ANSYS Workbench........................................................................ 121
4.6 Summary..................................................................................................................... 138
4.7 Review of Learning Objectives................................................................................ 138
Problems................................................................................................................................. 138
Contents
vii
5. Modeling and Solution Techniques................................................................................ 143
5.1 Introduction................................................................................................................ 143
5.2 Symmetry.................................................................................................................... 143
5.2.1 An Example................................................................................................... 144
5.3 Substructures (Superelements)................................................................................ 145
5.4 Equation Solving........................................................................................................ 146
5.4.1 Direct Methods (Gauss Elimination)......................................................... 146
5.4.2 Iterative Methods.......................................................................................... 146
5.4.3 An Example: Gauss Elimination................................................................. 146
5.4.4 An Example: Iterative Method.................................................................... 147
5.5 Nature of Finite Element Solutions......................................................................... 148
5.6 Convergence of FEA Solutions................................................................................. 149
5.7 Adaptivity (h-, p-, and hp-Methods)......................................................................... 149
5.8 Case Study with ANSYS Workbench........................................................................ 150
5.9 Summary..................................................................................................................... 161
5.10 Review of Learning Objectives................................................................................ 162
Problems................................................................................................................................. 162
6. Plate and Shell Analyses.................................................................................................... 165
6.1 Introduction................................................................................................................ 165
6.2 Review of Plate Theory............................................................................................. 165
6.2.1 Force and Stress Relations in Plates........................................................... 165
6.2.2 Thin Plate Theory (Kirchhoff Plate Theory)............................................. 167
6.2.2.1 Example: A Thin Plate.................................................................. 169
6.2.3 Thick Plate Theory (Mindlin Plate Theory).............................................. 170
6.2.4 Shell Theory................................................................................................... 171
6.2.4.1 Shell Example: A Cylindrical Container.................................... 171
6.3 Modeling of Plates and Shells.................................................................................. 172
6.4 Formulation of the Plate and Shell Elements......................................................... 173
6.4.1 Kirchhoff Plate Elements............................................................................. 173
6.4.2 Mindlin Plate Elements................................................................................ 174
6.4.3 Discrete Kirchhoff Elements....................................................................... 175
6.4.4 Flat Shell Elements........................................................................................ 175
6.4.5 Curved Shell Elements................................................................................. 176
6.5 Case Studies with ANSYS Workbench...................................................................... 177
6.6 Summary..................................................................................................................... 185
6.7 Review of Learning Objectives................................................................................ 185
Problems................................................................................................................................. 185
7. Three-Dimensional Elasticity........................................................................................... 189
7.1 Introduction................................................................................................................ 189
7.2 Review of Theory of Elasticity................................................................................. 189
7.2.1 Stress–Strain Relation.................................................................................. 190
7.2.2 Displacement................................................................................................. 191
7.2.3 Strain–Displacement Relation..................................................................... 191
7.2.4 Equilibrium Equations................................................................................. 191
7.2.5 Boundary Conditions................................................................................... 192
viii
Contents
7.2.6 Stress Analysis.............................................................................................. 192
Modeling of 3-D Elastic Structures......................................................................... 192
7.3.1 Mesh Discretization...................................................................................... 193
7.3.2 Boundary Conditions: Supports................................................................. 193
7.3.3 Boundary Conditions: Loads...................................................................... 194
7.3.4 Assembly Analysis: Contacts...................................................................... 194
7.4 Formulation of Solid Elements................................................................................. 195
7.4.1 General Formulation.................................................................................... 195
7.4.2 Typical Solid Element Types....................................................................... 196
7.4.3 Formulation of a Linear Hexahedral Element Type................................ 197
7.4.4 Treatment of Distributed Loads.................................................................. 200
7.5 Case Studies with ANSYS Workbench...................................................................... 200
7.6 Summary..................................................................................................................... 220
7.7 Review of Learning Objectives................................................................................ 220
Problems................................................................................................................................. 220
7.3
8. Structural Vibration and Dynamics................................................................................225
8.1 Introduction................................................................................................................225
8.2 Review of Basic Equations........................................................................................225
8.2.1 A Single DOF System................................................................................... 226
8.2.2 A Multi-DOF System.................................................................................... 228
8.2.2.1 Mass Matrices................................................................................ 228
8.2.2.2 Damping......................................................................................... 230
8.3 Formulation for Modal Analysis............................................................................. 231
8.3.1 Modal Equations........................................................................................... 233
8.4 Formulation for Frequency Response Analysis.................................................... 235
8.4.1 Modal Method............................................................................................... 235
8.4.2 Direct Method............................................................................................... 236
8.5 Formulation for Transient Response Analysis...................................................... 236
8.5.1 Direct Methods (Direct Integration Methods).......................................... 237
8.5.2 Modal Method............................................................................................... 238
8.6 Modeling Examples................................................................................................... 239
8.6.1 Modal Analysis............................................................................................. 239
8.6.2 Frequency Response Analysis.................................................................... 240
8.6.3 Transient Response Analysis...................................................................... 240
8.6.4 Cautions in Dynamic Analysis................................................................... 240
8.7 Case Studies with ANSYS Workbench...................................................................... 241
8.8 Summary..................................................................................................................... 260
8.9 Review of Learning Objectives................................................................................ 260
Problems................................................................................................................................. 260
9. Thermal Analysis................................................................................................................ 267
9.1 Introduction................................................................................................................ 267
9.2 Review of Basic Equations........................................................................................ 267
9.2.1 Thermal Analysis.......................................................................................... 267
9.2.1.1 Finite Element Formulation for Heat Conduction.................... 269
9.2.2 Thermal Stress Analysis.............................................................................. 269
9.2.2.1 1-D Case.......................................................................................... 270
9.2.2.2 2-D Cases........................................................................................ 271
Contents
ix
9.2.2.3 3-D Case.......................................................................................... 271
9.2.2.4 Notes on FEA for Thermal Stress Analysis............................... 271
9.3 Modeling of Thermal Problems............................................................................... 272
9.3.1 Thermal Analysis.......................................................................................... 272
9.3.2 Thermal Stress Analysis.............................................................................. 272
9.4 Case Studies with ANSYS Workbench...................................................................... 274
9.5 Summary..................................................................................................................... 292
9.6 Review of Learning Objectives................................................................................ 292
Problems................................................................................................................................. 293
10. Introduction to Fluid Analysis......................................................................................... 299
10.1 Introduction................................................................................................................ 299
10.2 Review of Basic Equations........................................................................................ 299
10.2.1 Describing Fluid Motion.............................................................................. 299
10.2.2 Types of Fluid Flow...................................................................................... 299
10.2.3 Navier–Stokes Equations.............................................................................300
10.3 Modeling of Fluid Flow............................................................................................. 301
10.3.1 Fluid Domain................................................................................................ 301
10.3.2 Meshing.......................................................................................................... 301
10.3.3 Boundary Conditions................................................................................... 301
10.3.4 Solution Visualization.................................................................................. 302
10.4 Case Studies with ANSYS Workbench...................................................................... 303
10.5 Summary..................................................................................................................... 325
10.6 Review of Learning Objectives................................................................................ 325
Problems................................................................................................................................. 326
11. Design Optimization.......................................................................................................... 331
11.1 Introduction................................................................................................................ 331
11.2 Topology Optimization............................................................................................. 331
11.3 Parametric Optimization.......................................................................................... 332
11.4 Design Space Exploration for Parametric Optimization...................................... 332
11.4.1 Design of Experiments................................................................................. 333
11.4.2 Response Surface Optimization................................................................. 335
11.5 Case Studies with ANSYS Workbench...................................................................... 335
11.6 Summary..................................................................................................................... 355
11.7 Review of Learning Objectives................................................................................ 355
Problems................................................................................................................................. 356
12. Failure Analysis................................................................................................................... 361
12.1 Introduction................................................................................................................ 361
12.2 Static Failure............................................................................................................... 361
12.2.1 Ductile Failure............................................................................................... 361
12.2.1.1 Maximum Shear Stress Theory (Tresca Criterion)................... 361
12.2.1.2 Distortion Energy Theory (von Mises Criterion)...................... 362
12.2.2 Brittle Failure................................................................................................. 362
12.2.2.1 Maximum Normal Stress Theory............................................... 362
12.2.2.2 Mohr–Coulomb Theory............................................................... 362
12.3 Fatigue Failure............................................................................................................ 363
x
Contents
12.3.1 Soderberg Failure Criterion.........................................................................364
12.3.2 Goodman Failure Criterion.........................................................................364
12.3.3 Gerber Failure Criterion............................................................................... 365
12.4 Buckling Failure......................................................................................................... 366
12.5 Case Studies with ANSYS Workbench...................................................................... 367
12.6 Summary..................................................................................................................... 376
12.7 Review of Learning Objectives................................................................................ 377
Problems................................................................................................................................. 377
Appendix 1: Review of Matrix Algebra................................................................................. 381
Appendix 2: Photo Credits........................................................................................................ 387
References.................................................................................................................................... 389
Preface
This book presents an integrated introduction to the basic theories of finite element
analysis (FEA) and the use of ANSYS® Workbench in the modeling and simulation of
engineering problems. It is intended for undergraduate students in engineering and
for engineers who have no previous experience with the FEA and wish to learn the
basic theory and software usage from a single volume. The book covers the basic concepts in the FEA using simple mechanics problems as examples, and leads to discussions and applications of the one-dimensional bar and beam elements, two-dimensional
plane stress and plane strain elements, plate and shell elements, and three-dimensional
solid elements in the analyses of structural stresses, vibrations and dynamics, thermal
responses, fluid flows, optimizations, and failures. In each of these application areas,
ANSYS Workbench is introduced through hands-on case studies. Clear understanding of
the FEA principles, element behaviors, solution procedures, and correct usage of the FEA
software is emphasized throughout the book.
The materials in this book are good for teaching undergraduate FEA courses in engineering to students who need a general background in the FEA so that they can apply FEA in
their design and analysis of structures using ANSYS Workbench or other FEA software.
Presentations of FEA theories and formulations in the book are kept to a minimum, aimed
solely for readers to understand the behaviors of the various elements and FEA solutions.
Emphases are placed on the techniques in FEA modeling and simulation, such as in building a quality FEA mesh, verifying the results, and employing the symmetry of a structure. The book includes 12 chapters and can be used for an undergraduate FEA course in
one semester (15 weeks with three 50-minute sessions each week) or in two quarters (10
weeks per quarter with three 50-minute sessions each week). It is recommended that computer lab sessions on using ANSYS Workbench be arranged for each chapter. Case studies
included in the book provide step-by-step instructions on solving a variety of problems of
moderate size and level of sophistication in ANSYS Workbench, and can be covered in the
lab sessions. Modeling tips are available in the case studies wherever applicable to give
students an immediate opportunity to use the skills they learn in a problem-solving context. Homework problems and projects using ANSYS Workbench software are provided at
the end of each chapter.
The Solutions Manual contains solutions to all end-of-chapter problems. The manual
features a compact presentation of solutions for over 100 problems. Each end-of-chapter
problem that requires the use of an FEA tool is solved in ANSYS Workbench. For such
a problem, the Workbench solution, which contains crucial properties such as geometry,
material, mesh, boundary conditions, and simulation results, is saved using a Workbench
project file (.wbpj) as well as a project folder that contains the supporting files for the
Workbench project. The Solutions Manual, along with the Workbench solution files for
over 80 end-of-chapter simulation problems, is available to instructors from CRC Press.
Microsoft PowerPoint slides for each chapter are provided to instructors using this book.
In addition, Workbench files for the case studies are available to instructors for use in lectures or computer lab sessions. The additional material is available from the CRC website:
http://www.crcpress.com/product/isbn/9781439873847.
The materials in the book have been developed by the authors for undergraduate
courses on FEA at the Washington State University Vancouver and at the University
xi
xii
Preface
of Cincinnati for the last 10 years or so. The authors would like to thank many of their
­current and former students for their help in developing these lecture materials and
their contributions to many of the examples used in the book, and in particular we thank
students Qikai Xie, Fan Yang, and Bin He for their help in developing the supplementary
materials for this book. We are grateful for the students we met in our classes at Vancouver
and Cincinnati. This book could never have been written without their collective interest, feedback, and enthusiasm. Our special thanks go to the editorial staff at CRC Press,
Jonathan Plant, the Senior Editor, and Amber Donley, the Project Coordinator, for their
patience, cooperation, and professionalism in the production of this book. Much gratitude also goes to the reviewers for their valuable comments and suggestions. Of course,
we take full responsibility for any errors or mistakes in the book. We also ­welcome any
suggestions from readers that could help us improve the book in the future.
Xiaolin Chen
Vancouver, Washington
Yijun Liu
Cincinnati, Ohio
Authors
Dr. Xiaolin Chen is an associate professor of mechanical engineering at the Washington
State University Vancouver. She earned her BS in engineering mechanics from Shanghai
Jiao Tong University, MS in mechanical design and theory from the State Key Laboratory
of Mechanical System and Vibration affiliated with Shanghai Jiao Tong University, and
her PhD in mechanical engineering from the University of Cincinnati. Her research interests include computational mechanics, finite element analysis, boundary element analysis,
multiphysics phenomena and coupled-field problems, inverse problems, and model order
reduction techniques.
Dr. Yijun Liu is a professor of mechanical engineering at the University of Cincinnati.
He earned his BS and MS in aerospace engineering from Northwestern Polytechnical
University (China), and his PhD in theoretical and applied mechanics from the University
of Illinois at Urbana-Champaign. Prior to joining the faculty, he conducted postdoctoral
research at the Center of Nondestructive Evaluation of Iowa State University and worked
at Ford Motor Company as a CAE analyst. Dr. Liu’s interests are in computational mechanics, finite element method, boundary element method, and fast multipole method in modeling composite materials, fracture, fatigue, structural dynamics, and acoustics problems.
He is the author of the book Fast Multipole Boundary Element Method (Cambridge University
Press, 2009) and has had more than 100 papers published in technical journals and conference proceedings in areas of computational mechanics.
xiii
1
Introduction
1.1 Some Basic Concepts
The finite element method (FEM), or finite element analysis (FEA), is based on the idea of
building a complicated object with simple blocks, or, dividing a complicated object into
smaller and manageable pieces. Application of this simple idea can be found everywhere
in everyday life (Figure 1.1), as well as in engineering. For example, children play with
LEGO® toys by using many small pieces, each of very simple geometry, to build various objects such as trains, ships, or buildings. With more and more smaller pieces, these
objects will look more realistic.
In mathematical terms, this is simply the use of the limit concept, that is, to approach or
represent a smooth object with a finite number of simple pieces and increasing the number
of such pieces in order to improve the accuracy of this representation.
1.1.1 Why FEA?
Computers have revolutionized the practice of engineering. Design of a product that used
to be done by tedious hand drawings has been replaced by computer-aided design (CAD)
using computer graphics. Analysis of a design used to be done by hand calculations and
many of the testing have been replaced by computer simulations using computer-aided
engineering (CAE) software. Together, CAD, CAE, and computer-aided manufacturing
(CAM) have dramatically changed the landscape of engineering (Figure 1.2). For example,
a car, that used to take five to six years from design to product, can now be produced starting from the concept design to the manufacturing within a year using the CAD/CAE/
CAM technologies.
Among all the computational tools for CAE, the FEM is the most widely applied method
or one of the most powerful modern “calculators” available for engineering students and
professionals. FEA provides a way of virtually testing a product design. It helps users
understand their designs and implement appropriate design changes early in the product
development process. The adoption of FEA in the design cycle is driven by market pressure since it brings many benefits that will help companies make better products with
reduced development costs and time-to-market.
1.1.2 Finite Element Applications in Engineering
The FEM can be applied in solving the mathematical models of many engineering problems, from stress analysis of truss and frame structures or complicated machines, to
dynamic responses of automobiles, trains, or airplanes under different mechanical,
1
2
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 1.1
Objects built with simple and small pieces: (a) a fire engine built with LEGO® (http://lego.wikia.com/
wiki/10020_Santa_Fe_Super_Chief); and (b) a house built with many elements—bricks, beams, columns, panels, and so on.
CAD
Design
concept
CAE
Design
modeling
Design
analysis
Design
optimization
Customer
market
Design
evaluation
Packaging
Production
planning
Quality
control
Process
planning
CAM
FIGURE 1.2
A sketch of the computer-aided product development process.
thermal, or electromagnetic loading. There are numerous finite element applications
in industries, ranging from automotive, aerospace, defense, consumer products, and
industrial equipment to energy, transportation and construction, as shown by some
examples in Table 1.1. The applications of the FEA have also been extended to materials
science, biomedical engineering, geophysics, and many other emerging fields in recent
years.
TABLE 1.1
Examples of Engineering Applications Using FEA
Field of Study
Structural and solid mechanics
Heat transfer
Fluid flow
Electrostatics/electromagnetics
Examples of Engineering Applications
Offshore structure reliability analysis, vehicle crash simulation,
nuclear reactor component integrity analysis, wind turbine blade
design optimization
Electronics cooling modeling, casting modeling, combustion engine
heat-transfer analysis
Aerodynamic analysis of race car designs, modeling of airflow
patterns in buildings, seepage analysis through porous media
Field calculations in sensors and actuators, performance prediction of
antenna designs, electromagnetic interference suppression analysis
Introduction
3
FIGURE 1.3
Examples of FEA using ANSYS Workbench: (a) wind load simulation of an offshore platform (Courtesy of
ANSYS, Inc., http://www.ansys.com/Industries/Energy/Oil+&+Gas); (b) modal response of a steel frame
building with concrete slab floors (http://www.isvr.co.uk/modelling/); (c) underhood flow and thermal management (Courtesy of ANSYS, Inc., http://www.ansys.com/Industries/Automotive/Application+Highlights/
Underhood); and (d) electric field pattern of antenna mounted on helicopter (Courtesy of ANSYS, Inc., http://
www.ansys.com/Industries/Electronics+&+Semiconductor/Defense+&+Aerospace + Electronics).
1.1.3 FEA with ANSYS Workbench
Over the last few decades, many commercial programs have become available for conducting the FEA. Among a comprehensive range of finite element simulation solutions
­provided by leading CAE companies, ANSYS® Workbench is a user-friendly platform
designed to seamlessly integrate ANSYS, Inc.’s suite of advanced engineering simulation
technology. It offers bidirectional connection to major CAD systems. The Workbench environment is geared toward improving productivity and ease of use among engineering
teams. It has evolved as an indispensible tool for product development at a growing number of companies, finding applications in many diverse engineering fields (Figure 1.3).
1.1.4 A Brief History of FEA
An account of the historical development of FEM and the computational mechanics in
general was given by O. C. Zienkiewicz recently, which can be found in Reference [1]. The
foundation of the FEM was first developed by Courant in the early 1940s. The stiffness
method, a prelude of the FEM, was developed by Turner, Clough et al., in 1956. The name
“finite element” was coined by Clough in 1960. Computer implementation of FEM programs emerged during the early 1970s. To date, FEM has become one of the most widely
used and versatile analysis techniques. A few major milestones are as follows:
1943—Courant (Variational methods which laid the foundation for FEM)
1956—Turner, Clough, Martin, and Topp (Stiffness method)
1960—Clough (Coined “Finite Element,” solved plane problems)
1970s—Applications on “mainframe” computers
4
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 1.4
(a) A plate with a hole (CAD model); and (b) A FEM discretization (mesh).
1980s—Microcomputers, development of pre- and postprocessors (GUI)
1990s—Analysis of large structural systems, nonlinear, and dynamic problems
2000s—Analysis of multiphysics and multiscale problems
1.1.5 A General Procedure for FEA
To conduct an FEA, the following procedure is required in general:
• Divide the CAD/geometric model into pieces to create a “mesh” (a collection of
elements with nodes, Figure 1.4).
• Describe the behavior of the physical quantities on each element.
• Connect (assemble) the elements at the nodes to form an approximate system of
equations for the entire model.
• Apply loads and boundary conditions (e.g., to prevent the model from moving).
• Solve the system of equations involving unknown quantities at the nodes (e.g., the
displacements).
• Calculate the desired quantities (e.g., strains and stresses) at elements or nodes.
In commercial FEA software, this procedure is typically rearranged into the following
phases:
• Preprocessing (build FEM models, define element properties, and apply loads and
constraints)
• FEA solver (assemble and solve the FEM system of equations, calculate element
results)
• Postprocessing (sort and display the results)
1.2 An Example in FEA: Spring System
A glimpse into the steps involved in an FEA is provided through a simple example in this
section. We will look at a spring element and a spring system to gain insight into the basic
concepts of the FEM.
5
Introduction
1.2.1 One Spring Element
For the single element shown in Figure 1.5, we have:
Two nodes
Nodal displacements
Nodal forces
Spring constant (stiffness)
i, j
ui, uj (m, mm)
fi, fj (Newton)
k (N/m, N/mm)
Relationship between spring force F and elongation Δ is shown in Figure 1.6.
In the linear portion of the curve shown in Figure 1.6, we have
F = kΔ,
with Δ = uj − ui
(1.1)
where k = F/Δ(>0) is the stiffness of the spring (the force needed to produce a unit stretch).
Consider the equilibrium of forces for the spring. At node i, we have
fi = −F = k(uj − ui) = kui − kuj
fi
i
F
and at node j
fj = F = k(uj − ui) = kui − kuj
F
j
fj
In matrix form,
− k ui f i
=
k u j f j
k
−k
x
i
fi
ui
j
uj
k
fj
FIGURE 1.5
One spring element.
Linear
F
Nonlinear
k
∆
FIGURE 1.6
Force–displacement relation in a spring.
(1.2)
6
Finite Element Modeling and Simulation with ANSYS Workbench
or,
ku = f
(1.3)
where
k = element stiffness matrix
u = element nodal displacement vector
f = element nodal force vector
From the derivation, we see that the first equation in Equation 1.2 represents the equilibrium of forces at node i, while the second equation in Equation 1.2 represents the equilibrium of forces at node j. Note also that k is symmetric. Is k singular or nonsingular? That
is, can we solve the equation in Equation 1.2? If not, why?
1.2.2 A Spring System
For a system of multiple spring elements, we first write down the stiffness equation for
each spring and then “assemble” them together to form the stiffness equation for the whole
system. For example, for the two-spring system shown in Figure 1.7, we proceed as follows:
For element 1, we have
k1
−k
1
− k1 u1 f11
=
k1 u2 f 21
(1.4)
k2
−k
2
− k 2 u2 f12
=
k 2 u3 f 22
(1.5)
and for element 2,
where f im is the (internal) force acting on local node i of element m (i = 1, 2).
1.2.2.1 Assembly of Element Equations: Direct Approach
Consider the equilibrium of forces at node 1,
F1 = f11
F1
k1
1
u1, F1
FIGURE 1.7
A system of two spring elements.
1
x
f11
k2
2
3
u2, F2
u3, F3
7
Introduction
at node 2,
f21
f12
2
F2
F2 = f 21 + f12
and, at node 3,
F3 = f 22
f22
3
F3
Using Equations 1.4 and 1.5, we obtain
F1 = k1u1 − k1u2
F2 = − k1u1 + (k1 + k 2 )u2 − k 2u3
F3 = − k 2u2 + k 2u3
In matrix form, we have
k1
− k1
0
− k1
k1 + k 2
− k2
0 u1 F1
− k 2 u2 = F2
k 2 u3 F3
(1.6)
or
Ku = F
in which, K is the stiffness matrix (structure matrix) for the entire spring system.
1.2.2.1.1 An Alternative Way of Assembling the Whole Stiffness Matrix
“Enlarging” the stiffness matrices for elements 1 and 2, we have
k1
− k1
0
− k1
k1
0
0 u1 f11
0 u2 = f 21
0 u3 0
and
0
0
0
0
k2
− k2
0 u1 0
− k 2 u2 = f12
k 2 u3 f 22
(1.7)
8
Finite Element Modeling and Simulation with ANSYS Workbench
Adding the two matrix equations (i.e., using superposition), we have
− k1
k1 + k 2
− k2
k1
− k1
0
0 u1 f11
− k 2 u2 = f 21 + f12
k 2 u3 f 22
This is the same equation we derived by using the concept of equilibrium of forces.
1.2.2.2 Assembly of Element Equations: Energy Approach
We can also obtain the result using an energy method, for example, the principle of minimum potential energy. In fact, the energy approach is more general and considered the
foundation of the FEM. To proceed, we consider the strain energy U stored in the spring
system shown in Figure 1.5.
U=
1
1
1
1
k1∆ 12 + k 2∆ 2 2 = ∆ 1T k1∆ 1 + ∆ 2T k 2∆ 2
2
2
2
2
However,
∆ 1 = u2 − u1 = [−1
u1
1] ,
u2
∆ 2 = u3 − u2 = [−1
u2
1]
u3
We have
U=
1
[u1
2
1
= [u1
2
k1
u2 ]
− k1
u2
− k 2 u2
= (enlarging …)
k 2 u3
− k1 u1 1
+ [u2
k1 u2 2
k1
u3 ] − k1
0
− k1
k1 + k 2
− k2
k2
u3 ]
− k2
0 u1
− k 2 u2
k 2 u3
(1.8)
The potential of the external forces is
Ω = − F1u1 − F2u2 − F3u3 = −[u1
u2
F1
u3 ] F2
F
3
(1.9)
Thus, the total potential energy of the system is
1
Π = U + Ω = [u1
2
u2
k1
u3 ] − k1
0
− k1
k1 + k 2
− k2
0 u1
− k 2 u2 − [u1
k 2 u3
u2
F1
u3 ] F2
F
3
(1.10)
9
Introduction
which is a function of the three nodal displacements (u1,u2,u3). According to the principle
of minimum potential energy, for a system to be in equilibrium, the total potential energy
must be minimum, that is, dΠ = 0, or equivalently,
∂Π
= 0,
∂u1
∂Π
= 0,
∂u2
∂Π
= 0,
∂u3
(1.11)
which yield the same three equations as in Equation 1.6.
1.2.3 Boundary and Load Conditions
Assuming that node 1 is fixed, and same force P is applied at node 2 and node 3, that is
u1 = 0
and F2 = F3 = P
we have from Equation 1.6
k1
− k1
0
− k1
k1 + k 2
− k2
0 0 F1
− k 2 u2 = P
k 2 u3 P
which reduces to
k1 + k 2
−k
2
− k 2 u2 P
=
k 2 u3 P
and
F1 = −k1u2
Unknowns are
u2
u=
u3
and the reaction force F1 (if desired).
Solving the equations, we obtain the displacements
2P/k1
u2
=
u3 2P/k1 + P/k 2
and the reaction force
F1 = −2P
10
Finite Element Modeling and Simulation with ANSYS Workbench
1.2.4 Solution Verification
It is very important in FEA to verify the results you obtained through either hand calculations or analytical solutions in the literature. The following is a list of items to check based
on common sense or intuition, or analytical solutions if they are available.
• Deformed shape of the structure
• Equilibrium of the external forces (Reaction forces should balance with the applied
loads.)
• Order of magnitudes of the obtained values
Notes about the Spring Elements:
• Spring elements are only suitable for stiffness analysis.
• They are not suitable for stress analysis of the spring itself.
• There are spring elements with stiffness in the lateral direction, spring elements
for torsion, and so on.
1.2.5 Example Problems
EXAMPLE 1.1
k1
1
k2
P
2
3
k3
4
x
Given: For the spring system shown above,
k1 = 100 N/mm, k 2 = 200 N/mm, k 3 = 100 N/mm
P = 500 N, u1 = u4 = 0
Find:
a. The global stiffness matrix
b. Displacements of nodes 2 and 3
c. The reaction forces at nodes 1 and 4
d. The force in the spring 2
Solution
a. The element stiffness matrices are (make sure to put proper unit after each
number)
100
k1 =
−100
−100
(N/mm)
100
200
k2 =
−200
−200
(N/mm)
200
11
Introduction
100
k3 =
−100
−100
(N/mm)
100
Applying the superposition concept, we obtain the global stiffness matrix
for the spring system
u1
100
−100
K=
0
0
u2
u3
u4
−100
100 + 200
−200
0
0
−200
200 + 100
−100
0
0
−100
100
or
100
−100
K=
0
0
−100
300
−200
0
0
−200
300
−100
0
0
−100
100
which is symmetric and banded.
Equilibrium (FE) equation for the whole system is
100
−100
0
0
−100
300
−200
0
0
−200
300
−100
0 u1 F1
0 u2 F2
=
−100 u3 F3
100 u4 F4
b. Applying the BCs u1 = u4 = 0, F2 = 0, and F3 = P, and “deleting” the first and
fourth rows and columns, we have
300
−200
−200 u2 0
=
300 u3 P
Solving this equation, we obtain
u2 P/250 2
= (mm)
=
u3 3P/500 3
c. From the first and fourth equations in the system of FE equations, we obtain
the reaction forces
F1 = −100u2 = −200 (N)
F4 = −100u3 = −300 (N)
12
Finite Element Modeling and Simulation with ANSYS Workbench
d. The FE equation for spring (element) 2 is
−200 ui f i
=
200 u j f j
200
−200
Here i = 2, j = 3 for element 2. Thus we can calculate the spring force as
u2
200] = [−200
u3
F = f j = − f i = [−200
2
200] = 200 (N)
3
Check the results:
Draw the free-body diagram (FBD) of the system and consider the equilibrium of the forces.
F1 = –200 N
F4 = –300 N
P = 500 N
Equilibrium of the forces is satisfied!
EXAMPLE 1.2
4
k4
1
k1
4
F1
2
1
x
k2
F2
k3
2
3
3
5
Problem
For the spring system with arbitrarily numbered nodes and elements, as shown above,
find the global stiffness matrix.
Solution
First, we construct the following element connectivity table:
Element
1
2
3
4
Node i (1)
Node j (2)
4
2
3
2
2
3
5
1
This table specifies the global node numbers corresponding to the local node numbers
for each element.
13
Introduction
Then we write the element stiffness matrix for each element
u4
u2
k1
k1 =
− k1
u2
− k1
,
k1
k2
k2 =
−k2
−k2
,
k 2
u3
u3
u5
k3
k3 =
− k3
u2
− k3
,
k 3
k4
k4 =
−k4
−k4
k 4
u1
Finally, applying the superposition method, we obtain the global stiffness matrix as
follows:
u1
k4
−k
4
K= 0
0
0
u2
u3
u4
u5
−k4
k1 + k 2 + k 4
− k2
− k1
0
0
− k2
k2 + k3
0
− k3
0
− k1
0
k1
0
0
0
− k3
0
k 3
The matrix is symmetric, banded, but singular, as it should be.
After introducing the basic concepts, the section below introduces you to one of the
general-purpose finite element software tools—ANSYS Workbench.
1.3 Overview of ANSYS Workbench
ANSYS Workbench is a simulation platform that enables users to model and solve a wide
range of engineering problems using the FEA. It provides access to the ANSYS family
of design and analysis modules in an integrated simulation environment. This section
gives a brief overview of the different elements in the ANSYS Workbench simulation environment or the graphical-user interface (GUI). Readers are referred to ANSYS Workbench
user’s guide [2] for more detailed information.
1.3.1 The User Interface
The Workbench interface is composed primarily of a Toolbox region and a Project Schematic
region (Figure 1.8). The main use of the two regions is described next.
14
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 1.8
ANSYS Workbench user interface.
1.3.2 The Toolbox
The Toolbox contains the following four groups of systems:
Analysis Systems: Predefined analysis templates to be used to build your project,
including static structural, steady-state thermal, transient thermal, fluid flow,
modal, shape optimization, linear buckling, and many others.
Component Systems: Component applications that can be used to build or expand an
analysis system, including geometry import, engineering data, mesh, postprocessing, and others.
Custom Systems: Coupled-field analysis systems such as fluid solid interaction, prestress modal, thermal-stress, and others.
Design Exploration: Parametric optimization studies such as response surface optimization, parameters correlation, six sigma analysis, and others.
1.3.3 The Project Schematic
A project schematic, that is, a graphical representation of the workflow, can be built by
dragging predefined analysis templates or other components from the Toolbox and dropping them into the Project Schematic window. “Drag” here means to move the mouse while
holding down the left mouse button, and “drop” means to release the mouse button.
To build a project for static structural analysis, for instance, drag the Static Structural
template from the Toolbox and drop it into the rectangular box that appears in the Project
Schematic window. A standalone analysis system that contains the components needed for
static structural analysis is added to the project schematic as shown in Figure 1.9a. The
system consists of seven individual components called cells.
Alternatively, a standalone analysis can be created by double-clicking. For example,
double-click the Steady-State Thermal template from the Toolbox, and an independent
Introduction
15
FIGURE 1.9
Defining standalone analysis systems in the project schematic: (a) a standalone system; (b) two independent
standalone systems; (c) moving a system in a top-bottom configuration; and (d) moving a system in a side-byside configuration.
Steady-State Thermal system will be placed in the default location below the existing Static
Structural system, as shown in Figure 1.9b.
A system can be moved around another system in the project schematic. To move a system, click on the header cell (i.e., the cell titled Steady-State Thermal for the thermal system)
and drag it to a new place. Once you drag the header cell, dashed rectangles appear for the
possible new locations to drop the system. This is illustrated in Figures 1.9c and d for two
systems with initial top–bottom and side-by-side configurations, respectively.
To delete a system, click on the down arrow button at the upper left corner of the system
from the Project Schematic window, and then choose Delete from the drop-down context
menu.
In some cases, a project may contain two or more analysis systems that share data. For
example, a downstream modal analysis may use the same material, geometry, and model
data from the preceding structural analysis. To build such a project, create a standalone
system for Static Structural analysis. Then, drag the Modal analysis template from the
Toolbox and drop it onto the Model cell of the Static Structural system. Immediately before
the subsequent system is dropped, bounding boxes will appear on the Engineering Data,
Geometry, and Model cells of the first system, as shown in Figure 1.10a. After the system
is released, a project including two linked systems is created, as shown in Figure 1.10b,
where the linked cells indicate data sharing at the Model and above levels.
16
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 1.10
Defining linked analysis systems in the project schematic: (a) dropping the second (subsequent) system onto
the Model cell of the first system to share data at the model and above levels; (b) two systems that are linked.
1.3.4 Working with Cells
Cells are components that make up an analysis system. You may launch an application by
double-clicking a cell. To initiate an action other than the default action, right-click on a
cell to view its context menu options. The following list comprises the types of cells available in ANSYS Workbench and their intended functions:
Engineering Data: Define or edit material models to be used in an analysis.
Geometry: Create, import, or edit the geometry model used for analysis.
Model/Mesh: Assign material, define coordinate system, and generate mesh for the
model.
Setup: Apply loads, boundary conditions, and configure the analysis settings.
Solution: Access the model solution or share solution data with other downstream
systems.
Results: Indicate the results availability and status (also referred to as postprocessing).
As the data flows through a system, a cell’s state can quickly change. ANSYS Workbench
provides a state indicator icon placed on the right side of the cell. Table 1.2 describes the
indicator icons and the various cell states available in ANSYS Workbench. For more information, please refer to ANSYS Workbench user’s guide [2].
1.3.5 The Menu Bar
The menu bar is the horizontal bar anchored at the top of the Workbench user interface. It
provides access to the following functions:
File Menu: Create a new project, open an existing project, save the current project,
and so on.
View Menu: Control the window/workspace layout, customize the toolbox, and so on.
Tools Menu: Update the project and set the license preferences and other user options.
Units Menu: Select the unit system and specify unit display options.
Help Menu: Get help for ANSYS Workbench.
17
Introduction
TABLE 1.2
Indicator Icons and Descriptions of the Various Cell States
Cell State
Indicator
Description
Unfulfilled
Need upstream data to proceed
Refresh required
A refresh action is needed as a result of changes made on
upstream data
Attention required
User interaction with the cell is needed to proceed
Update required
An update action is needed as a result of changes made on
upstream data
Up to date
Data are up to date and no attention is required
Input changes pending
An update or refresh action is needed to recalculate based
on changes made to upstream cells
Interrupted
Solution has been interrupted. A resume or update action
will make the solver continue from the interrupted point
Pending
Solution is in progress
Source: Courtesy of ANSYS Workbench User’s Guide, Release 14.5, ANSYS, Inc., 2012.
In the chapters that follow, the use of ANSYS Workbench will be presented in a step-bystep fashion in the context of real-world problem solving that may involve different modeling concepts or different physics. It is worth noting that, although a commercial finite
element program enables you to deal with a wide range of engineering problems with
complex geometry, constraints, and material behaviors, it is your responsibility to understand the underlying physics to be able to setup a problem correctly. Your FEA results, no
matter how pretty they may look, are only as good as the assumptions and decisions you
have made in building your model.
1.4 Summary
In this chapter, the basic concepts in the FEM are introduced. The spring system is used
as an example to show how to establish the element stiffness matrices, to assemble
the finite element equations for a system from element stiffness matrices, and to solve
the FE equations. Verifying the FE results is emphasized. ANSYS Workbench environment is briefly introduced. The concepts and procedures introduced in this chapter are
very simple and yet very important for studying the finite element analyses of other
problems.
18
Finite Element Modeling and Simulation with ANSYS Workbench
PROBLEMS
1.1 Answer the following questions briefly:
a. What is the physical meaning of the FE equations (for either an element or the
whole structure)?
b. What is the procedure in using the FEM?
1.2 Answer the following questions briefly:
a. What is a finite element discretization or a finite element mesh?
b. Define the terms: node and element.
1.3 Using ANSYS Workbench, solve the following problems:
a. Create a project schematic that includes structural analysis of two design
parts of the same vehicle.
b. Create a project schematic that allows sequential thermal-stress coupling
analysis.
1.4 Consider the spring system shown below. Find the global stiffness matrix for the
entire system.
1
k3
k1
3
k5
4
k4
k2
2
5
F5
k6
6
x
1.5 Consider the spring system shown below. Find the global stiffness matrix for the
entire system.
k2
1
k1
3
k3
2
x
4
k4
k6
5
F5
6
k5
1.6 A spring system is shown below
1
k1
F3
k2
2
k3
3
4
F4
x
Given: k1 = 80 N/mm, k2 = 100 N/mm, k3 = 160 N/mm, F3 = 200 N, F4 = 100 N, and
nodes 1 and 2 are fixed;
Find:
a. Global stiffness matrix
b. Displacements of nodes 3 and 4
19
Introduction
c. Reaction forces at nodes 1 and 2
d. Forces in springs 1 and 2
1.7 A spring system is shown below
1
2
k1
k2
3
F3
k4
4
k5
5
k3
x
Given: k1 = k2 = k3 = k4 = k5 = k, F3 = P, and nodes 1, 2, and 5 are fixed;
Find:
a. Global stiffness matrix
b. Displacements of nodes 3 and 4
c. Reaction forces at nodes 1, 2, and 5
d. Forces in springs 2, 4, and 5
2
Bars and Trusses
2.1 Introduction
This chapter introduces you to the simplest one-dimensional (1-D) structural element,
namely the bar element, and the FEA of truss structures using such element. Trusses
are commonly used in the design of buildings, bridges, and towers (Figure 2.1). They are
triangulated frameworks composed of slender bars whose ends are connected through
bolts, pins, rivets, and so on. Truss structures create large, open, and uninterrupted space,
and offer lightweight and economical solutions to many engineering situations. If a truss,
along with the applied load, lies in a single plane, it is called a planar truss. If it has members and joints extending into the three-dimensional (3-D) space, it is then a space truss.
Most structural analysis problems such as stress and strain analysis can be treated as
linear static problems, based on the following assumptions:
1. Small deformations (loading pattern is not changed due to the deformed shape)
2. Elastic materials (no plasticity or failures)
3. Static loads (the load is applied to the structure in a slow or steady fashion)
Linear analysis can provide most of the information about the behavior of a structure,
and can be a good approximation for many analyses. It is also the basis of nonlinear FEA
in most of the cases. In Chapters 2 through 7, only linear static responses of structures are
considered.
2.2 Review of the 1-D Elasticity Theory
We begin by examining the problem of an axially loaded bar based on 1-D linear elasticity.
Consider a uniform prismatic bar shown in Figure 2.2. The parameters L, A, and E are the
length, cross-sectional area, and elastic modulus of the bar, respectively.
Let u, ε, and σ be the displacement, strain, and stress, respectively (all in the axial direction and functions of x only), we have the following basic relations:
Strain–displacement relation:
ε( x) =
du( x)
dx
(2.1)
21
22
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 2.1
Truss examples: (a) Montreal Biosphere Museum (http://en.wikipedia.org/wiki/Montreal_Biosph%C3%A8re).
(b) Betsy Ross Bridge (http://en.wikipedia.org/wiki/Betsy_Ross_Bridge).
A, E
x
P
L
FIGURE 2.2
An axially loaded elastic bar.
Stress–strain relation:
σ( x) = Eε( x)
(2.2)
d σ( x )
+ f ( x) = 0
dx
(2.3)
Equilibrium equation:
where f(x) is the body force (force per unit volume, such as gravitational and magnetic
forces) inside the bar. To obtain the displacement, strain, and stress field in a bar, Equations
2.1 through 2.3 need to be solved under given boundary conditions, which can be done
readily for a single bar, but can be tedious for a network of bars or a truss structure made
of many bars.
2.3 Modeling of Trusses
For the truss analysis, it is often assumed that: (1) the bar members are of uniform cross
sections and are joined together by frictionless pins, and (2) loads are applied to joints only
and not in between joints along the truss members. It is based on these assumptions that
the truss members are considered to carry only axial loads and have negligible bending
resistance.
23
Bars and Trusses
(a)
Roof
Purlin
Gusset
(b)
FIGURE 2.3
Modeling of a planar roof truss: (a) Physical structure. (b) Discrete model.
Given the fact that a truss is an assembly of axial bars, it is generally satisfactory to treat
each truss member as a discrete element—bar element. A bar element is a 1-D finite element to describe the deformation, strain, and stress in a slender structure member that
often has a uniform cross section and is loaded in tension or compression only along its
axis.
Let us take a planar roof structure as an example to illustrate the truss model idealization process. In an actual structure pictured in Figure 2.3a, purlins are used to support
the roof and they further transmit the load to gusset plates that connect the members.
In the idealized model shown in Figure 2.3b, each truss member is simply replaced by a
two-node bar element. A joint condition is modeled by node sharing between elements
connected end-to-end, and loads are imposed only on nodes. Components such as purlins
and gussets are neglected to avoid construction details that do not contribute significantly
to the overall truss load–deflection behavior. Other factors such as the weight of the truss
members can be considered negligible, compared to the loads they carry. If it were to be
included, a member’s weight can always be applied to its two ends, half the weight at each
end node. Experience has shown that the deformation and stress calculations for such
idealized models are often sufficient for the overall design and analysis purposes of truss
structures [3].
2.4 Formulation of the Bar Element
In this section, we will formulate the equations for the bar element based on the 1-D elasticity theory. The structural behavior of the element, or the element stiffness matrix, can be
established using two approaches: the direct approach and the energy approach. The two
methods are discussed in the following.
2.4.1 Stiffness Matrix: Direct Method
A bar element with two end nodes is presented in Figure 2.4.
24
Finite Element Modeling and Simulation with ANSYS Workbench
uj
ui
fi
i
x
j
A, E
fj
L
FIGURE 2.4
Notation for a bar element.
Assuming that the displacement u is varying linearly along the axis of the bar, that is, in
terms of the two nodal values ui and uj, we can write
x
x
u( x) = 1 − ui + u j
L
L
(2.4)
We have
ε =
u j − ui
∆
=
L
L
(∆ = elongation)
(2.5)
E∆
L
(2.6)
( F = force in bar)
(2.7)
σ = Eε =
We also have
F
A
σ =
Thus, Equations 2.6 and 2.7 lead to
F=
EA
∆ = k∆
L
where k = EA/L is the stiffness of the bar. That is, the bar behaves like a spring in this case
and we conclude that the element stiffness matrix is
k
k=
−k
EA
−k L
=
k EA
−
L
EA
L
EA
L
−
or
k=
EA 1
L −1
−1
1
(2.8)
25
Bars and Trusses
This can be verified by considering the equilibrium of the forces at the two nodes.
The Element equilibrium equation is
−1 ui f i
=
1 u j f j
EA 1
L −1
(2.9)
Degree of Freedom (DOF): Number of components of the displacement vector at a node.
For 1-D bar element along the x-axis, we have one DOF at each node.
Physical Meaning of the Coefficients in k: The jth column of k (here j = 1 or 2) represents the
forces applied to the bar to maintain a deformed shape with unit displacement at node j
and zero displacement at the other node.
2.4.2 Stiffness Matrix: Energy Approach
We derive the same stiffness matrix for the bar using a formal approach which can be
applied to many other more complicated situations.
First, we define two linear shape functions as follows (Figure 2.5):
N i (ξ ) = 1 − ξ , N j (ξ ) = ξ
(2.10)
where
ξ =
x
, 0≤ ξ≤1
L
(2.11)
From Equation 2.4, we can write the displacement as
u( x) = u(ξ) = N i (ξ)ui + N j (ξ)u j
or
ui
N j = Nu
u j
u = N i
(2.12)
Strain is given by Equations 2.1 and 2.12 as
ε =
du d
=
N u = Bu
dx dx
Nj(ξ )
Ni(ξ )
1
i
ξ=0
FIGURE 2.5
The shape functions for a bar element.
(2.13)
j
ξ=1
ξ
i
ξ=0
1
j
ξ=1
ξ
26
Finite Element Modeling and Simulation with ANSYS Workbench
where B is the element strain–displacement matrix, which is
B=
d
N i (ξ )
dx
N j (ξ) =
d
N i (ξ )
dξ
B = −1/L
1/L
N j (ξ) ⋅
dξ
dx
that is,
(2.14)
Stress can be written as
σ = E ε = EBu
(2.15)
Consider the strain energy stored in the bar
U=
1
1
σ T ε dV =
(u TBT EBu) dV
2
2
∫
∫
V
=
V
1 T
u (BT EB) dV u
2
V
∫
(2.16)
where Equations 2.13 and 2.15 have been used.
The potential of the external forces is written as (this is by definition, and remember the
negative sign)
Ω = − f iui − f ju j = −u T f
(2.17)
The total potential of the system is
Π=U+Ω
which yields by using Equations 2.16 and 2.17
Π=
1 T
u (BT EB) dV u − u T f
2
V
∫
(2.18)
Setting dΠ = 0 by the principle of minimum potential energy, we obtain (verify this)
BT EB dV u = f
V
∫(
)
or
ku = f
(2.19)
27
Bars and Trusses
where
k=
∫ (B EB) dV
T
(2.20)
V
is the element stiffness matrix.
Equation 2.20 is a general result which can be used for the construction of other types
of elements.
Now, we evaluate Equation 2.20 for the bar element by using Equation 2.14
L
k=
−1/L
E −1/L
1/L
0
∫
1/L Adx =
−1
1
EA 1
L −1
which is the same as we derived earlier using the direct method.
Note that from Equations 2.16 and 2.20, the strain energy in the element can be written as
U=
1 T
u ku
2
(2.21)
In the future, once we obtain an expression like Equation 2.16, we can immediately recognize that the matrix in between the displacement vectors is the stiffness matrix. Recall
that for a spring, the strain energy can be written as
U=
1 2 1 T
k∆ = ∆ k∆
2
2
Thus, result (2.21) goes back to the simple spring case again.
2.4.3 Treatment of Distributed Load
Distributed axial load q (N/mm, N/m, lb/in) (Figure 2.6) can be converted into two
­equivalent nodal forces using the shape functions. Consider the work done by the distributed load q,
L
1
1
Wq =
u( x)q( x)dx =
2
2
∫
0
L
∫
0
L
Ni ( x)
u j
q( x)dx
N j ( x )
0
1
(Nu) q(x)dx = ui
2
∫
T
L
=
1 T
u NT q( x)dx
2
∫
(2.22)
0
q
i
fi
x
j
q
fj
i
FIGURE 2.6
Conversion of a distributed load on one element.
j
q
28
Finite Element Modeling and Simulation with ANSYS Workbench
q
1
3
2
qL/2
qL/2
qL
1
2
3
FIGURE 2.7
Conversion of a distributed load with constant intensity q on two elements.
The work done by the equivalent nodal forces are
W fq =
1 q
1
1
f i ui + f jqu j = uT fq
2
2
2
(2.23)
Setting Wq = W fq and using Equations 2.22 and 2.23, we obtain the equivalent nodal force
vector
f i q
fq = q =
f j
L
∫
L
NT q( x)dx =
0
Ni ( x)
N ( x) q( x)dx
j
0
∫
(2.24)
which is valid for any distributions of q. For example, if q is a constant, we have
L
qL/2
1 − x/L
fq = q
dx = qL/2
/
x
L
0
∫
(2.25)
that is, equivalent nodal forces can be added to replace the distributed load as shown in
Figure 2.7.
2.4.4 Bar Element in 2-D and 3-D
To analyze the truss structures in 2-D or 3-D, we need to extend the 1-D bar element formulation to 2-D or 3-D. In the following, we take a look at the formulation for the 2-D case.
2.4.4.1 2-D Case
Local
Global
x, y
ui′, vi′
X, Y
ui, vi
1 DOF at each node
2 DOFs at each node
Note that lateral displacement vi′ does not contribute to the stretch of the bar within the
linear theory (Figure 2.8). Displacement vectors in the local and global coordinates are
related as follows:
29
Bars and Trusses
y
x
j
Y
ui′
i
ui
θ
vi
X
FIGURE 2.8
Local and global coordinates for a bar in 2-D space.
ui
m
vi
ui
vii = −ui sin θ + vi cos θ = − m l
vi
ui′ = ui cos θ + vi sin θ = l
where l = cos θ, m = sin θ.
In matrix form,
′
ui l
′ =
vi − m
m ui
l vi
(2.26)
or,
i
u′i = Tu
where the transformation matrix
m
l
l
T =
−m
(2.27)
is orthogonal, that is, T −1 = T T .
For the two nodes of the bar element, we have
ui′ l
′
vi − m
′ =
u j 0
v′j 0
m
l
0
0
0
0
l
−m
0 ui
0 vi
m u j
l v j
(2.28)
0
T
(2.29)
or,
T
u′ = Tu with T =
0
30
Finite Element Modeling and Simulation with ANSYS Workbench
The nodal forces are transformed in the same way,
f′ = Tf
(2.30)
In the local coordinate system, we have
−1 ui′ f i′
=
1 u′j f j′
EA 1
L −1
Augmenting this equation, we write
1
EA 0
L −1
0
−1
0
1
0
0
0
0
0
0 ui′ f i′
0 vi′ 0
=
0 u′j f j′
0 v′j 0
or,
k′ u′ = f′
Using transformations given in Equations 2.29 and 2.30, we obtain
k′ Tu = Tf
Multiplying both sides by T T and noticing that T T T = I, we obtain
T T k′ Tu = f
(2.31)
Thus, the element stiffness matrix k in the global coordinate system is
k = T T k′ T
(2.32)
which is a 4 × 4 symmetric matrix.
Explicit form is
ui
l
EA lm
k=
L −l 2
−lm
2
vi
lm
m2
−lm
− m2
uj
2
−l
−lm
l2
lm
vj
−lm
− m2
lm
m2
Calculation of the directional cosines l and m:
l = cos θ =
Yj − Yi
X j − Xi
, m = sin θ =
L
L
(2.33)
31
Bars and Trusses
The structure stiffness matrix is assembled by using the element stiffness matrices in the
usual way as in the 1-D case.
2.4.4.2 3-D Case
Local
Global
x, y, z
ui′, vi′, wi′
X, Y, Z
ui, vi, wi
1 DOF at each node
3 DOFs at each node
Similar to the 2-D case, element stiffness matrices in the 3-D case are calculated in the
local coordinate systems first and then transformed into the global coordinate system (X,
Y, and Z) where they are assembled (Figure 2.9). The transformation relation is
ui′ lX
′
vi = mX
w ′ n
i X
lZ ui
mZ vi
nZ wi
lY
mY
nY
(2.34)
where (lX,lY,lZ), (mX,mY,mZ), and (nX,nY,nZ) are the direction cosines of the local x, y, and z
coordinate axis in the global coordinate system, respectively. FEM software packages will
do this transformation automatically.
Therefore, the input data for bar elements are simply:
• Coordinates (X, Y, Z) for each node
• E and A for each element (length L can be computed from the coordinates of the
two nodes)
2.4.5 Element Stress
Once the nodal displacement is obtained for an element, the stress within the element can
be calculated using the basic relations. For example, for 2-D cases, we proceed as follows:
u
σ = Eε = EB
u
′
i
′
j
1
= E −
L
1l
L 0
y
x
Y
i
X
Z
FIGURE 2.9
Local and global coordinates for a bar in 3-D space.
z
0
l
m
0
j
ui
0 vi
m u j
v j
32
Finite Element Modeling and Simulation with ANSYS Workbench
That is,
E
−l
L
σ =
−m
ui
v
i
m
u j
v j
l
(2.35)
which is a general formula for 2-D bar elements.
2.5 Examples with Bar Elements
We now look at examples of using the bar element in the analysis of 1-D stress problems
and plane truss problems.
EXAMPLE 2.1
2A, E
A, E
1
1
2
L
P
2
3
L
x
Problem
Find the stresses in the two-bar assembly which is loaded with force P, and constrained
at the two ends, as shown in the above figure.
Solution
Use two 1-D bar elements.
For element 1,
k1 =
u1
u2
2EA 1
L −1
−1
1
For element 2,
k2 =
u2
u3
EA 1
L −1
−1
1
Imagine a frictionless pin at node 2, which connects the two elements. We can assemble the global FE equation as follows:
33
Bars and Trusses
2
EA
−2
L
0
−2
3
−1
0 u1 F1
−1 u2 = F2
1 u3 F3
Load and boundary conditions (BCs) are
u1 = u3 = 0, F2 = P
FE equation becomes
2
EA
−2
L
0
−2
3
−1
0 0 F1
−1 u2 = P
1 0 F3
“Deleting” the first row and column, and the third row and column, we obtain
EA
[3]{u2 } = {P}
L
Thus,
u2 =
PL
3EA
and
0
u1
PL
1
u2 =
u 3EA 0
3
Stress in element 1 is
σ 1 = Eε1 = EB1u1 = E −1/L
=E
u2 − u1
E PL
=
− 0
L
L 3EA
u1
1/L
u2
P
=
3A
Similarly, stress in element 2 is
u2
1/L
u3
P
u − u2
E
PL
=E 3
= 0 −
=−
3EA
3A
L
L
σ 2 = Eε 2 = EB2u 2 = E −1/L
which indicates that bar 2 is in compression.
34
Finite Element Modeling and Simulation with ANSYS Workbench
Check the results: Draw the FBD as follows and check the equilibrium of the
structures.
–2P/3
–P/3
P
Notes:
• In this case, the calculated stresses in elements 1 and 2 are exact. It will not
help if we further divide element 1 or 2 into smaller elements.
• For tapered bars, averaged values of the cross-sectional areas should be used
for the elements.
• We need to find the displacements first in order to find the stresses, and this
approach is called the displacement-based FEM.
EXAMPLE 2.2
θ
A, E
1
1
2
L
P
2
3
x
L
Problem
Determine the support reaction forces at the two ends of the bar shown above, given
the following:
P = 6.0 × 10 4 N ,
E = 2.0 × 10 4 N/mm 2 ,
2
A = 250 mm , L = 150 mm,
∆ = 1.2 mm
Solution
We first check to see if contact of the bar with the wall on the right will occur or not. To
do this, we imagine the wall on the right is removed and calculate the displacement at
the right end
∆0 =
PL
(6.0 × 10 4 )(150)
=
= 1.8 mm > ∆ = 1.2 mm
EA (2.0 × 10 4 )(250)
Thus, contact occurs and the wall on the right should be accounted for in the analysis.
The global FE equation is found to be
1
EA
−1
L
0
−1
2
−1
0 u1 F1
−1 u2 = F2
1 u3 F3
35
Bars and Trusses
The load and boundary conditions are
F2 = P = 6.0 × 10 4 N
u1 = 0, u3 = ∆ = 1.2 mm
The FE equation becomes
−1
2
−1
1
EA
−1
L
0
0 0 F1
−1 u2 = P
1 ∆ F3
The second equation gives
EA
2
L
u2
−1 = {P}
∆
that is,
EA
EA
[2]{u2 } = P + L ∆
L
Solving this, we obtain
u2 =
1 PL
+ ∆ = 1.5 mm
2 EA
and
u1 0
u2 = 1.5 (mm)
u 1.2
3
To calculate the support reaction forces, we apply the first and third equations in the
global FE equation.
The first equation gives
F1 =
EA
1
L
−1
u1
EA
0 u2 =
(−u2 ) = −5.0 × 10 4 N
L
u
3
and the third equation gives,
F3 =
EA
0
L
−1
u1
EA
1 u2 =
(−u2 + u3 ) = −1.0 × 10 4 N
L
u
3
Check the results: Again, we can draw the free-body diagram to verify that the equilibrium of the forces is satisfied.
36
Finite Element Modeling and Simulation with ANSYS Workbench
EXAMPLE 2.3
3
45°
2
P2
2
Y
P1
1
45°
X
1
A simple plane truss is made of two identical bars (with E, A, and L), and loaded as
shown in the above figure.
Find
a. displacement of node 2;
b. stress in each bar.
Solution
This simple structure is used here to demonstrate the FEA procedure using the bar element in 2-D space. In local coordinate systems, we have
k1′ =
−1
= k ′2
1
EA 1
L −1
These two matrices cannot be assembled together, because they are in different coordinate systems. We need to convert them into global coordinate system OXY.
Element 1:
θ = 45o , l = m =
2
2
Using Equation 2.32 or 2.33, we obtain the stiffness matrix in the global system
u1
v1
u2
v2
1
EA 1
k1 = T1T k1′ T1 =
2L −1
−1
1
1
−1
−1
−1
−1
1
1
−1
−1
1
1
Element 2:
θ = 135o , l = −
2
2
, m=
2
2
u2
v2
u3
v3
1
EA −1
k 2 = T2T k ′2 T2 =
2L −1
1
−1
1
1
−1
−1
1
1
−1
1
−1
−1
1
37
Bars and Trusses
Assemble the structure FE equation,
u1
v1
u2
v2
u3
v3
1
1
EA −1
2L −1
0
0
1
1
−1
−1
0
0
−1
−1
2
0
−1
1
−1
−1
0
2
1
−1
0
0
−1
1
1
−1
0 u1 F1X
0 v1 F1Y
1 u2 F2 X
=
−1 v2 F2Y
−1 u3 F3 X
1 v3 F3Y
Load and boundary conditions (BCs):
u1 = v1 = u3 = v3 = 0, F2 X = P1 , F2Y = P2
Condensed FE equation,
EA 2
2L 0
0 u2 P1
=
2 v2 P2
Solving this, we obtain the displacement of node 2,
u2
L P1
=
v2 EA P2
Using Equation 2.34, we calculate the stresses in the two bars,
σ1 =
E 2
−1
L 2
E 2
σ2 =
1
L 2
−1
−1
1
0
L 0
2
1
(P1 + P2 )
=
EA P1 2 A
P2
−1
P1
L P2
2
1
(P1 − P2 )
=
0
EA 2 A
0
Check the results: Again, we need to check the equilibrium conditions, symmetry, antisymmetry, and so on, of the FEA results.
EXAMPLE 2.4: (MULTIPOINT CONSTRAINT)
y′
P
2
2
L
1
3
Y
3
1
45°
X
x′
38
Finite Element Modeling and Simulation with ANSYS Workbench
For the plane truss shown above,
P = 1000 kN,
L = 1 m,
E = 210 GPa,
A = 6.0 × 10 −4 m2 for elements 1 and 2,
A = 6 2 × 10 −4 m2 for element 3.
Determine the displacements and reaction forces.
Solution
We have an inclined roller at node 3, which needs special attention in the FE solution.
We first assemble the global FE equation for the truss.
Element 1:
θ = 90 o , l = 0, m = 1
u1
v1
u2
0
(210 × 10 )(6.0 × 10 ) 0
k1 =
0
1
0
0
1
0
−1
0
0
0
0
0
−1
(N/m)
0
1
v3
−4
9
v2
Element 2:
θ = 0o ,
l = 1, m = 0
u2
v2
u3
1
(210 × 10 )(6.0 × 10 ) 0
k2 =
−1
1
0
0
0
0
0
−1
0
1
0
9
−4
0
0
(N/m)
0
0
Element 3:
θ = 45o , l =
1
1
, m=
2
2
u1
0.5
(210 × 10 9 )(6 2 × 10 −4 ) 0.5
k3 =
−0.5
2
−0.5
v1
u3
0.5
0.5
−0.5
−0.5
−0.5
−0.5
0.5
0.5
v3
−0.5
−0.5
(N/m)
0.5
0.5
39
Bars and Trusses
The global FE equation is,
0.5
5
1260 × 10
Sym.
0.5
1.5
0
0
1
0
−1
0
1
−0.5
−0.5
−1
0
1.5
−0.5 u1 F1X
−0.5 v1 F1Y
0 u2 F2 X
=
0 v2 F2Y
0.5 u3 F3 X
0.5 v3 F3Y
Load and boundary conditions (BCs):
u1 = v1 = v2 = 0, and v3′ = 0,
F2 X = P , F3 x′ = 0.
From the transformation relation and the BCs, we have
2
ν′3 = −
2
2 u3
2
( − u3 + v 3 ) = 0 ,
=
2 v3
2
that is,
u3 − v3 = 0
This is a multipoint constraint (MPC).
Similarly, we have a relation for the force at node 3,
2
F3 x′ =
2
2 F3 X
2
( F3 X + F3Y ) = 0,
=
2 F3Y
2
that is,
F3X + F3Y = 0
Applying the load and BCs in the structure FE equation by “deleting” the first,
­second, and fourth rows and columns, we have
1
1260 × 10 5 −1
0
−1
1.5
0.5
0 u2 P
0.5 u3 = F3 X
0.5 v3 F3Y
Further, from the MPC and the force relation at node 3, the equation becomes,
1
1260 × 10 5 −1
0
−1
1.5
0.5
0 u2 P
0.5 u3 = F3 X
0.5 u3 − F3 X
which is
1
1260 × 10 −1
0
5
−1
P
u2
2 = F3 X
u
3
1 − F3 X
The third equation yields,
F3 X = −1260 × 10 5 u3
40
Finite Element Modeling and Simulation with ANSYS Workbench
Substituting this into the second equation and rearranging, we have
−1 u2 P
=
3 u3 0
1
1260 × 10 5
−1
Solving this, we obtain the displacements,
3P 0.01191
u2
1
=
(m )
=
5
u
2520
10
×
3
P 0.003968
From the global FE equation, we can calculate the reaction forces,
−0.5
−0.5
0
1.5
0.5
0
F1X
0
F
1Y
5
F2Y = 1260 × 10 0
F
−1
3X
0
F3Y
−0.5
−500
−0.5 u2 −500
0 u3 = 0.0 (kN)
0.5 v3
−500
0.5
500
Check the results!
A general MPC can be described as,
∑A u
j
j
=0
j
where Aj ’s are constants and uj ’s are nodal displacement components. In FE software,
users only need to specify this relation to the software. The software will take care of
the solution process.
2.6 Case Study with ANSYS Workbench
Problem Description: Truss bridges can span long distances and support heavy weights
without intermediate supports. They are economical to construct and are available in a
wide variety of styles. Consider the following planar truss, constructed of wooden timbers, which can be used in parallel to form bridges. Determine the deflections at each joint
of the truss under the given loading conditions.
Material: Douglas fir
1m
1m
1m
30 kN
1m
30 kN
1m
30 kN
E = 13.1 GPa
ν = 0.29
Member cross section:
Height = 6 cm
Width = 6 cm
Solution
To solve the problem with ANSYS® Workbench, we employ the following steps:
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Woodtruss.wbpj.”
41
Bars and Trusses
Step 2: Create a Static Structural (ANSYS) Analysis System
Drag the Static Structural (ANSYS) icon from the Analysis Systems Toolbox window
and drop it inside the highlighted green rectangle in the Project Schematic window to create a standalone static structural analysis system.
Step 3: Add a New Material
Double-click (or right-click and choose Edit) on the Engineering Data cell in the
above Project Schematic to edit or add a material. In the following Engineering
Data interface which replaces the Project Schematic, click the empty box highlighted below and type a name, for example, “Douglas Fir,” for the new material.
Type ‘Douglas Fir’ in
the empty box.
42
Finite Element Modeling and Simulation with ANSYS Workbench
Select “Douglas Fir” from the Outline window, and double-click Isotropic Elasticity
under Linear Elastic in the leftmost Toolbox window. Enter “1.31E10” for Young’s
Modulus and “0.29” for Poisson’s Ratio in the bottom center Properties window.
Click the Return to Project button to go back to the Project Schematic.
Step 4: Launch the DesignModeler Program
Ensure Line Bodies is checked in the Properties of Schematic A3: Geometry window.
Double-click the Geometry cell to launch DesignModeler, and select “Meter” as
length unit in the Units pop-up window.
43
Bars and Trusses
Step 5: Create Line Sketch
Click the Sketching tab and select Settings. Turn on Show in 2D and Snap under
Grid options. Use the default value of “5 m” for Major Grid Spacing and “5” for
Minor-Steps per Major.
Click a start point and then an end point in the Graphics window to draw a line.
Draw 13 lines as shown in the sketch below. After completion, click Generate to
create a line sketch.
Click the first point and
then click the second
point to draw a line.
13
12
5
1
6
7
2
8
9
3
10
11
4
Step 6: Create Line Body from Sketch
Check off the Grid options under Settings of Sketching Toolboxes. Switch to the
Modeling tab. Note that a new item named Sketch1 now appears underneath
XYPlane in the Tree Outline.
Select Lines from Sketches from the Concept drop-down menu.
44
Finite Element Modeling and Simulation with ANSYS Workbench
Select Sketch1 from the Tree Outline and click Apply to confirm on the Base Objects
selection in the Details of Line1. Click Generate to complete the line body
creation.
Step 7: Create a Cross Section
Select a Cross Section of Rectangular from the Concept drop-down menu. A new
item named Rect1 is now added underneath the Cross Section in the Tree Outline.
Bars and Trusses
In the Details of Rect1 under Dimensions, enter “0.06 m” for both B and H.
Step 8: Assign Cross Section to Line Body
Select the Line Body underneath 1Part, 1 Body in the Tree Outline. In the Details of
Line Body, assign Rect1 to the Cross Section selection. Click Close DesignModeler
to exit the program.
45
46
Finite Element Modeling and Simulation with ANSYS Workbench
Step 9: Launch the Static Structural (ANSYS) Program
Double-click the Model cell to launch the Static Structural (ANSYS) program. Note
that in the Details of “Line Body” the material is assigned to Structural Steel by
default. Click to the right of the Assignment field and select Douglas Fir from the
drop-down context menu.
Step 10: Generate Mesh
In the Details of “Mesh,” enter a fairly large number, say, “10 m,” for the Element
Size, to ensure each member is meshed with only one element. In the Outline of
Project, right-click on Mesh and select Generate Mesh.
47
Bars and Trusses
Step 11: Apply Boundary Conditions
In the Outline of Project, right-click on Static Structural (A5) and select Insert and
then Fixed Support. After completion, a Fixed Support item is added underneath
Static Structural (A5) in the project outline tree.
Select the two points as shown below in the Graphics window. In the Details of
“Fixed Support,” click Apply to confirm on the Geometry selection. After completion, a Fixed Support boundary condition will be added to the selected two
points.
Turn on Vertex filter.
Ctrl-click points to
make multiple
selections.
1
2
48
Finite Element Modeling and Simulation with ANSYS Workbench
Step 12: Apply Loads
In the Outline of Project, right-click on Static Structural (A5) and select Insert and
then Force.
Select the three points as shown below in the Graphics window. In the Details of
“Force,” click Apply to confirm on the Geometry selection. Also underneath
the Details, change the Define By selection to Components and enter “-90000N”
for the Y Component. A downward red arrow will appear on the selected three
points in the Graphics window.
Turn on Vertex filter.
Ctrl-click points to
make multiple
selections.
1
2
3
Alternatively, the load can be applied to each of the three points individually
by inserting Force three times under Static Structural (A5). In this case, enter
“-30000N” for the Y Component of each individual Force item.
Bars and Trusses
Step 13: Retrieve Solution
Insert a Total Deformation item by right-clicking on Solution (A6) in the Outline tree.
Right-click on Solution (A6) in the Outline tree and select Solve. The program will start
to solve the model.
After completion, click Total Deformation in the Outline to review the total deformation
results.
49
50
Finite Element Modeling and Simulation with ANSYS Workbench
Modeling tips: To get the reaction force, a Force Reaction probe can be inserted by
right-clicking on Solution (A6) in the Outline tree as shown below.
In the Details of “Force Reaction,” select the Fixed Support as the Boundary Condition.
Bars and Trusses
Right-click on Solution (A6) in the Outline tree and select Evaluate All Results.
After completion, click Force Reaction in the Outline to review results.
Note here that the reaction force is found to be 90,000 N in the positive Y-direction.
This is because a boundary condition has been applied earlier to the two fixed ends in
one step (see Step 11). To avoid summing of the force reaction, two fixed conditions can
be inserted instead in Step 11, one for each end. The reaction forces at an individual
support can then be displayed by selecting the support of interest from the drop-down
menu of Boundary Condition in the Details of “Force Reaction.”
51
52
Finite Element Modeling and Simulation with ANSYS Workbench
2.7 Summary
In this chapter, we studied the bar elements which can be used in truss analysis. The concept of the shape functions is introduced and the derivations of the stiffness matrices using
the energy approach are introduced. Treatment of distributed loads is discussed and several examples are studied. A planar truss structure is analyzed using ANSYS Workbench.
It provides basic modeling techniques and shows step-by-step how Workbench can be used
to determine the deformation and reaction forces in trusses.
2.8 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1. Set up simplified finite element models for truss structures.
2. Derive the stiffness matrix for plane bar elements using direct and energy
approaches.
3. Explain the concept of shape functions and their characteristics for bar elements.
4. Find the equivalent nodal loads of distributed forces on bars.
5. Determine the displacement and stress of a truss using hand calculation to verify
the finite element solutions.
6. Apply the general bar element stiffness matrix to the analysis of simple trusses.
7. Create line sketches and new material definition in ANSYS Workbench.
8. Perform static structural analyses on trusses using ANSYS Workbench.
PROBLEMS
2.1 A bar assembly is loaded with force P at one end and constrained at the other
end, as shown in the figure below. Determine:
a. The displacement at node 2 and node 3
b. The stress in the bar assembly
E, 2A
E, A
1
2
L
E, A
3
L
P
4
L
53
Bars and Trusses
2.2 A simple structure with the profile shown below is loaded with force P, and
­constrained at one end. Young’s modulus E is the same for all the bars. The
cross-sectional areas are shown in the figure. Use 1-D bar elements to approximate the deformation and the stresses in the structure.
A
3A
A
3A
P
A
A
5L
2L
5L
2L
2.3 The plane truss is loaded with force P as shown below. Constants E and A for
each bar are as shown in the diagram. Determine:
a. The nodal displacements
b. The reaction forces
c. The stresses in bar elements
Y
L
L
1
3
L
4
E, A
E, A
E, A
2
X
P
2.4 The plane truss is loaded with force P as shown below. Constants E and A for
each bar are as shown in the diagram. Determine:
a. The nodal displacements
b. The reaction forces
c. The stresses in bar elements
L
Y
L
4
3
E, A
E, 2 2 A
45°
45°
1
P
E, A
2
X
54
Finite Element Modeling and Simulation with ANSYS Workbench
2.5 The plane truss is loaded with force P as shown below. Constants E and A for
each bar are as shown in the diagram. Determine:
a. The nodal displacements
b. The stresses in bar elements
Y
3
P
E, 2 2A
45°
L
E, A
1
45°
2
X
E, A
2.6 The plane truss is supported as shown below. Young’s modulus E is the same
for all the bars. The cross-sectional areas are shown in the figure. Suppose that
the node 2 settles by an amount of δ as shown. Determine the stresses in each
bar element using the FEM.
3
Y
2 2A
2
A
1
L
A
δ
L
X
2.7 The plane truss is loaded with force P as shown below. Young’s modulus E is
the same for all the bars. The cross-sectional areas are shown in the figure.
Determine:
a. The nodal displacements
b. The stresses in each bar element
Y
3
L
A
A
1
4
2 2A
2 2A
P
A
L
2
X
L
55
Bars and Trusses
2.8 The plane truss is loaded with force P as shown below. Young’s modulus E is
the same for all the bars. The cross-sectional areas are shown in the figure.
Determine the nodal displacements and the reaction forces.
Y
4
L/2
L/2
2A
5A
5
5A
1
2A
P
5A
A
2A
2
L
3
X
L
2.9 The roof truss shown below is made of Douglas fir timbers of a 5 mm × 5 mm
cross section. Use ANSYS Workbench to determine the truss deformation and the
support reaction forces.
1 kN
1 kN
1 kN
0.25 m
1 kN
1 kN
0.25 m
0.2 m
0.2 m
0.25 m
0.25 m
0.25 m
0.25 m
0.2 m
2.10 The truss tower crane shown below is made of structural steel rectangular bars
of a 3 mm × 3 mm cross section. Use ANSYS Workbench to determine the truss
deformation and the support reaction forces.
0.1 m
0.1 m
0.1 m
0.2 m
0.5 m
6 kN
0.5 m
0.5 m
0.5 m
0.5 m
0.5 m
0.5 m
0.5 m
56
Finite Element Modeling and Simulation with ANSYS Workbench
2.11 The truss bridge shown below is made of Douglas fir timbers of a 4 mm × 4 mm
cross section. Use ANSYS Workbench to determine the truss deformation and the
support reactions.
0.15 m
0.25 m
0.5 m
1.5 kN
0.5 m
1.5 kN
0.5 m
1.5 kN
0.5 m
1.5 kN
0.5 m
1.5 kN
0.5 m
0.5 m
2.12 The truss transmission tower shown below is made of structural steel members of a 2.5 mm × 2.5 mm cross section. Use ANSYS Workbench to determine
the truss deformation and the support reactions.
0.6 m
1 kN
0.8 m
1 kN
0.8 m
1.5 kN
1.5 kN
1.2 m
1.8 m
0.8 m 0.5 m 0.3 m 1 m 0.3 m 0.5 m 0.8 m
3
Beams and Frames
3.1 Introduction
Beams and frames are frequently used in constructions, in engineering equipment, and in
everyday life, for example, in buildings, lifting equipment, vehicles, and exercise machines
(Figure 3.1).
Beams are slender structural members subjected primarily to transverse loads. A beam
is geometrically similar to a bar in that its longitudinal dimension is significantly larger
than the two transverse dimensions. Unlike bars, the deformation in a beam is predominantly bending in transverse directions. Such a bending-dominated deformation is the
primary mechanism for a beam to resist transverse loads. In this chapter, we will use
the term “general beam” for a beam that is subjected to both bending and axial forces, and
the term “simple beam” for a beam subjected to only bending forces. The term “frame” is
used for structures constructed of two or more rigidly connected beams.
3.2 Review of the Beam Theory
We start with a brief review of the simple beam theory. Essential features of the two wellknown beam models, the Euler–Bernoulli beam and the Timoshenko beam, will be introduced, followed by discussions on the stress, strain, and deflection relations in simple
beam theory.
3.2.1 Euler–Bernoulli Beam and Timoshenko Beam
Euler–Bernoulli beam and Timoshenko beam, as shown in Figure 3.2, are two common
models that are used in the structural analysis of beams and frames. Both models have at
their core the assumption of small deformation and linear elastic isotropic material behavior. They are applicable to beams with uniform cross sections.
For a Euler–Bernoulli beam, it is assumed that the forces on a beam only cause the beam
to bend. There is no transverse shear deformation occurring in the beam bending. The
neutral axis, an axis that passes through the centroid of each beam cross section, does not
change in length after the deformation. A planar cross section perpendicular to the neutral axis remains plane and perpendicular to the neutral axis after deformation. Due to its
neglect of shear strain effects, the Euler–Bernoulli model tends to slightly ­underestimate
57
58
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 3.1
Examples of beams and frames: (a) a car frame (http://www.carbasics-1950.com/); and (b) an exercise machine.
(From http://www.nibbledaily.com/body-by-jake-cardio-cruiser/.)
Reference:
Deformed:
Euler–Bernoulli beam
Timoshenko beam
FIGURE 3.2
Two common beam models.
the beam deformation. The model is considered suitable for describing the behavior of
thin or long beams.
The Timoshenko beam model accounts for both transverse shear deformation and bending deformation. In this model, a planar cross section remains plane, but does not remain
normal to the neutral axis after deformation. By taking into account shear strain effects,
the Timoshenko model is physically more realistic for describing the behavior of thick or
short beams.
To determine the deflection of a laterally loaded beam of the same material and uniform
cross section, we may view the beam as a 1-D model with the transverse load assigned to
its neutral axis. As illustrated in Figure 3.3, for a Euler–Bernoulli beam, we need only one
variable—the vertical displacement (v), to describe the beam deflection at any point along
v
90°
v
90°
dv
θ=
dx
Euler–Bernoulli beam
FIGURE 3.3
Deflection of Euler–Bernoulli beam and Timoshenko beam.
θ
dv
dx
Timoshenko beam
59
Beams and Frames
the neutral axis. For a Timoshenko beam, both the vertical displacement (v) and the rotation angle of beam cross section (θ) are needed for a complete description of its deformed
configuration.
3.2.2 Stress, Strain, Deflection, and Their Relations
Simple beam bending is often analyzed using the Euler–Bernoulli beam theory, which is
popular in engineering owing to its simplicity and is referred to as the engineering beam
theory by many. To understand the stress state in simple beam bending, imagine that the
beam is made of many infinitesimally thin layers deposited on top of each other, as shown
in Figure 3.4.
Under the assumption that the neutral axis does not change in length, the material layer
located along the axis will stay stress free. However, the beam will bend concave upward
to resist a downward transverse load. Bending makes layers above the neutral axis axially
compressed and the ones underneath axially stretched. The further away a material layer
is from the neutral axis, the more it is strained. The axial strain induced by bending is
called the bending strain, which is related to the deflected beam curvature κ(x) based on
differential geometry as follows:
ε( x) = − yκ( x) = − y
d 2v
dx 2
(3.1)
where y is the vertical distance of a thin layer from the neutral axis and v the beam deflection.
The internal resisting bending moment M(x) is a function of the axial bending stress σ(x)
from the theory of elasticity. Together with Equation 3.1, we have:
M( x ) =
∫
∫
− y σ( x)dA = Ey 2 κ( x) dA = EI κ( x) = EI
d 2v
dx 2
(3.2)
where E is the material elastic modulus, I is the area moment of inertia of the beam crosssection with respect to the z-axis, and the product of E and I is called the bending stiffness.
Combining Equations 3.1 and 3.2, we arrive at the following flexure formula for the bending stress σ(x):
σ( x ) = −
M( x ) y
I
(3.3)
The above simple beam theory is analogous to the uniaxial Hooke’s law. As shown in
Table 3.1, the bending moment is linearly proportional to the deflected beam curvature
y
x
FIGURE 3.4
Bending produces axial stress σ(x) along thin material layers in a beam.
60
Finite Element Modeling and Simulation with ANSYS Workbench
TABLE 3.1
Analogy between the Constitutive Equations for Bars and Beams
Bar
Beam
Stress Measurement
Strain Measurement
Axial stress: σ(x)
Bending moment: M(x)
Axial strain: ε(x)
σ(x) = Eε(x)
d2v
Curvature: 2
dx
M( x) = EI
θ=
dv
dx
κ=
d2v
dx 2
dθ
dx
Curvature: κ(x)
Rotation: θ (x)
Deflection: v(x)
Constitutive Equation
M = EIκ
Distributed load: q(x)
q=
Shear force: Q(x)
dQ
dx
Bending moment: M(x)
Q=
dM
dx
FIGURE 3.5
The governing equations for a simple beam.
through a bending stiffness (EI) constant. This resembles the linear stress–strain relationship described by Hooke’s law. The basic equations that govern the problems of simple
beam bending are summarized in Figure 3.5. The equations included here will later be
used in the formulation of finite element equations for beams.
3.3 Modeling of Beams and Frames
Modeling is an idealization process. Engineers seek to simplify problems and model real
physical structures at an adequate level of detail in design and analysis to strike a balance
between efficiency and accuracy. Some considerations on cross sections, support conditions, and model simplification of beams and frames are discussed next.
3.3.1 Cross Sections and Strong/Weak Axis
Beams are available in various cross-sectional shapes. There are rectangular hollow tubes,
I-beams, C-beams, L-beams, T-beams, and W-beams, to name a few. Figure 3.6 illustrates
some common shapes for beam cross sections.
We have learned from Equation 3.2 that the bending stiffness (EI) measures a beam’s
ability to resist bending. The higher the bending stiffness, the less the beam is likely to
bend. Drawing on our everyday experience, it is more difficult to bend a flat ruler with its
flat (wide) surface facing forward rather than facing up, as shown in Figure 3.7. It is because
the moment of inertia (I) is a cross-sectional property sensitive to the distribution of material with respect to an axis.
61
Beams and Frames
FIGURE 3.6
Common beam cross-section profiles.
Large load capacity in its
strong-axis orientation
Small load capacity in its
weak-axis orientation
FIGURE 3.7
A beam loaded on its strong and weak axes.
To keep the deflection within acceptable limits, beams need to have sufficiently large
bending stiffness to withstand the applied load. It is important that we select beams of
efficient cross-section shapes and use beams on their strong-axes of bending, especially in
cases when switching to materials of high elastic modulus is deemed costly or infeasible
due to other constraints.
3.3.2 Support Conditions
There are many different ways of supporting beams. For example, a beam can be fixed
at both ends, or fixed at one end and pinned at the other end. Generally speaking, beams
have three types of end support conditions: fixed support, pinned support, and roller support. A fixed support is an anchor condition that has zero translation and zero rotation at
the supported end. A pinned support is a hinge condition that prevents any translation but
does not prevent rotation about the hinge axis. A roller support allows for both rotation
and translation along the surface on which the roller moves but prevents any translation
normal to that surface. Figure 3.8 illustrates various types of beams named based on their
support conditions. As shown, a simply supported beam is supported by a hinge at one
end and by a roller at the other end. A cantilever beam is fixed at one end and free at the
other end. A fixed-pinned beam is fixed at one end and pinned at the other end, while a
fixed end beam is one that is clamped at both ends.
In general, support conditions are important considerations in determining the mathematical solution of a physical problem, and inappropriate support is a common cause of
errors. When we model beams and frames, bear in mind that valid boundary conditions
need to reflect the underlying physical construction and be sufficient to prevent the structure from any rigid-body motion (both translation and rotation).
62
Finite Element Modeling and Simulation with ANSYS Workbench
Simply supported beam
Fixed-pinned beam
Cantilever beam
Fixed end beam
FIGURE 3.8
Beam supports and beam types.
FIGURE 3.9
A swing set and its simplified line model.
3.3.3 Conversion of a Physical Model into a Line Model
Aside from the support idealization, conceptual models are widely adopted in the analyses
of beams and frames to achieve modeling efficiency. The idea of model conceptualization
stems from the uniform cross-section assumption. Under the assumption, it is apparent
that a beam needs only to be modeled at the center axis (neutral axis) of the actual 3-D
beam structure, as shown in Figure 3.9 through a swing set example.
For beams and frames, the conceptual model is also known as the line model, which
consists of only lines or curves in general. After a line model is created, cross-sectional
properties and other data such as material properties, boundary conditions, and loads can
be specified for the analysis. Once a problem is fully defined, solutions can be obtained
readily after the line model is discretized into a line mesh using beam elements. In contrast to the truss modeling, where each truss member is discretized into a single bar element, many line segments (element divisions) are typically needed in a line mesh made
with beam elements in order to obtain more accurate results.
3.4 Formulation of the Beam Element
In this section, we discuss the finite element formulation based on the simple beam theory
(Euler–Bernoulli beam). The beam element stiffness matrix will be established using both
direct and energy methods.
63
Beams and Frames
y
vi , Fi
i
vj , Fj
j
E,I
θi , Mi
L
θ j , Mj
x
FIGURE 3.10
Notion for a simple beam element in 2-D.
The setup of a simple beam element in 2-D space is shown in Figure 3.10, where the
variables are:
L, I, and E = Length, moment of inertia of the cross-sectional area, and elastic modulus of the beam, respectively
length (L), moment of inertia of the cross-sectional area (I), and elastic modulus of the
beam (E), respectively
v = v(x) deflection (lateral displacement) of the neutral axis of the beam
θ = dv/dx rotation of the beam about the z-axis
Q = Q(x) (internal) shear force
M = M(x) (internal) bending moment about z-axis
Fi, Mi, Fj, Mj applied (external) lateral forces and moments at node i and j, respectively
For simplicity of presentation, we will restrict our attention to beam element formulation in 2-D space in the following discussion. For a beam element in 3-D space, the element
stiffness equation can be formed in the local (2-D) coordinate system first and then transformed into the global (3-D) coordinate system to be assembled.
3.4.1 Element Stiffness Equation: The Direct Approach
We first apply the direct method to establish the beam stiffness matrix using the results
from elementary beam theory. The FE equation for a beam takes the form
k11
k
21
k 31
k 41
k12
k 22
k 32
k 42
k13
k 23
k 33
k 43
k14 vi Fi
k 24 θi Mi
=
k 34 v j Fj
k 44 θj M j
(3.4)
Recall that each column in the stiffness matrix represents the forces needed to keep the
element in a special deformed shape. For example, the first column represents the forces/
moments to keep the shape with vi = 1, θi = vj = θj = 0 as shown in Figure 3.11a. Thus, using
the results from strength of materials for a cantilever beam with a force k11 and moment k21
applied at the free end, we have
vi =
k L2 k L
k11L3 k 21L2
−
= 1 and θi = − 11 + 21 = 0
3EI
2EI
2EI
EI
64
Finite Element Modeling and Simulation with ANSYS Workbench
(a)
1
(b)
0.15
0.1
0.5
0.05
0
0
(c)
1
0.5
0
1
N1(x)
0
0.5
1
N2(x)
(d)
0
–0.05
0.5
0
0
–0.1
0.5
1
–0.15
0
N3(x)
0.5
1
N4 (x)
FIGURE 3.11
Four configurations or shapes for the simple beam element.
Solving this system of equations, we obtain k11 and k21. Using the equilibrium conditions
of the beam, we obtain k31 and k41, and thus the first column of the stiffness matrix.
Using the same procedure in calculating columns 2, 3, and 4 of the matrix in Equation
3.4 (corresponding to shapes in Figures 3.11b through d), we obtain the element stiffness
equation as:
12
EI 6L
L3 −12
6L
6L
4L2
−6L
2L2
−12
−6L
12
−6L
6L vi Fi
2L2 θi Mi
=
−6L v j Fj
4L2 θj M j
(3.5)
3.4.2 Element Stiffness Equation: The Energy Approach
To derive the stiffness matrix in Equation 3.5 using the formal or energy approach, we
introduce four shape functions (as shown in Figure 3.11), which can be represented mathematically as
N1( x) = 1 − 3 x 2/L2 + 2x 3/L3
N 2 ( x) = x − 2x 2/L + x 3/L2
N 3 ( x) = 3 x 2/L2 − 2x 3/L3
N 4 ( x) = − x 2/L + x 3/L2
(3.6)
65
Beams and Frames
Then, on each element, we can represent the deflection of the beam (v) using shape functions and the corresponding nodal values as
v( x) = Nu = [N1( x)
N 2 ( x)
vi
θ
i
N 4 ( x)]
v j
θ j
N 3 ( x)
(3.7)
which is a cubic function. Note that,
N1 + N 3 = 1
N2 + N3L + N4 = x
which implies that the rigid-body motion is represented correctly by the assumed deformed
shape of the beam.
To derive the beam element stiffness matrix, we consider the curvature of the beam,
which is
d 2v
d2
=
Nu = Bu
dx 2
dx 2
(3.8)
where the strain–displacement matrix B is given by
d2
N = N1′′( x) N 2′′( x)
dx 2
4 6x
6 12x
= − 2 + 3
− + 2
L
L L
L
B=
N 4′′( x)
N 3′′( x)
2 6x
− + 2
L L
6 12x
− 3
L2
L
Strain energy stored in the beam element is
U=
1
σ T ε dV
2
∫
V
Applying the basic equations in the simple beam theory, we have
1
U=
2
L
∫∫
0 A
L
=
T
L
1
1
My 1 My
M T Mdx
−
−
dAdx =
2
I
E
I
EI
∫
0
T
∫
2
0
∫
0
1 T T
u B EIB dx u
2
L
=
L
d v
1
1 d v
EI 2 dx =
(Bu)T EI (Bu) dx
2
2
2 dx
dx
2
∫
0
(3.9)
66
Finite Element Modeling and Simulation with ANSYS Workbench
We conclude that the stiffness matrix for the simple beam element is
L
∫ B EIB dx
k=
T
(3.10)
0
Applying the result in Equation 3.10 and carrying out the integration, we arrive at the
same stiffness matrix as given in Equation 3.5.
3.4.3 Treatment of Distributed Loads
To convert a distributed load into nodal forces and moments (Figure 3.12), we consider
again the work done by the distributed load q
L
1
1
Wq =
v( x)q( x)dx =
2
2
∫
0
L
∫
0
L
1
(Nu) q(x)dx = uT NT q(x)dx
2
∫
T
0
The work done by the equivalent nodal forces (and moments) is
W fq =
1
[ vi
2
θi
Fiq
q
Mi 1
θ j] q = uT fq
Fj 2
Mq
j
vj
By equating Wq = W fq , we obtain the equivalent nodal force vector as
L
fq =
∫ N q(x)dx
T
(3.11)
0
which is valid for arbitrary distributions of q(x). For constant q, we have the results shown
in Figure 3.13. An example of this result is given in Figure 3.14.
q(x)
i
L
Fjq
Fiq
Miq
i
x
j
j
Mjq
FIGURE 3.12
Conversion of the distributed lateral load into nodal forces and moments.
67
Beams and Frames
q
i
qL/2
qL/2
qL2/12
x
j
L
i
j
qL2/12
FIGURE 3.13
Conversion of a constant distributed lateral load into nodal forces and moments.
q
L
L
qL/2
qL
qL2/12
L
L
FIGURE 3.14
Conversion of a constant distributed lateral load on two beam elements.
3.4.4 Stiffness Matrix for a General Beam Element
Combining the axial stiffness (from the bar element), we further arrive at the stiffness
matrix for a general 2-D beam element as
vi
θi
0
0
12EI
L3
6EI
L2
6EI
L2
4 EI
L
ui
EA
L
0
0
k=
EA
−
L
0
0
0
−
12EI
L3
6EI
L2
0
6EI
L2
2EI
L
−
uj
−
vj
EA
L
0
0
EA
L
0
0
0
12EI
L3
6EI
− 2
L
−
0
12EI
L3
6EI
− 2
L
θj
6EI
L2
2EI
L
0
6EI
− 2
L
4EI
L
0
(3.12)
68
Finite Element Modeling and Simulation with ANSYS Workbench
3.5 Examples with Beam Elements
EXAMPLE 3.1
y
P
1
1
2
M
E, I
2
x
3
L
L
Given:
The beam shown above is clamped at the two ends and acted upon by the force P and
moment M in the midspan.
Find:
The deflection and rotation at the center node and the reaction forces and moments at
the two ends.
Solution
Element stiffness matrices are
v1
θ1
v2
θ2
12
6L
−12
6L
6L
4L2
−6L
2L2
−12
−6L
12
−6L
6L
2L2
−6L
4L2
v2
θ2
v3
θ3
12
EI 6L
k2 = 3
L −12
6L
6L
4L2
−6L
2L2
−12
−6L
12
−6L
6L
2L2
−6L
4L2
EI
k1 = 3
L
Global FE equation is
v1
θ1
v2
θ2
v3
12
6L
EI −12
L3 6L
0
0
6L
4L2
−6L
2L2
0
0
−12
−6L
24
0
−12
6L
6L
2L2
0
8L2
−6L
2L2
0
0
−12
−6L
12
−6L
θ3
0 v1 F1Y
0 θ1 M1
6L v2 F2Y
=
2L2 θ 2 M2
−6L v3 F3Y
4L2 θ 3 M3
Loads and constraints (BCs) are
F2Y = − P ,
M2 = M ,
v1 = v3 = θ1 = θ 3 = 0
69
Beams and Frames
Reduced FE equation
EI
L3
0 v2 − P
=
8L2 θ 2 M
24
0
Solving this, we obtain
2
v2
L − PL
=
θ 2 24EI 3 M
From the global FE equation, we obtain the reaction forces and moments
−12
F1Y
M
1 EI −6L
= 3
F3Y L −12
M3
6L
6L
2P + 3 M/L
2
2L v2 1 PL + M
=
−6L θ 2 4 2P − 3 M/L
2
− PL + M
2L
Stresses in the beam at the two ends can be calculated using the formula
σ = σx = −
My
I
Note that the FE solution is exact for this problem according to the simple beam
­t heory, since no distributed load is present between the nodes. Recall that
EI
d 4v
= q( x)
dx 4
If q(x) = 0, then exact solution for the deflection v is a cubic function of x, which is
exactly what described by the shape functions given in Equation 3.6.
EXAMPLE 3.2
y
1
p
L
E, I
2
x
Given:
A cantilever beam with distributed lateral load p as shown above.
Find:
The deflection and rotation at the right end, the reaction force and moment at the left
end.
70
Finite Element Modeling and Simulation with ANSYS Workbench
Solution
The work-equivalent nodal loads are shown below,
y
f
1
L
E, I
2
m
x
where
m = pL2/12
f = pL/2,
Applying the FE equation, we have
12
EI 6L
3
L −12
6L
6L
4L2
−6L
2L2
−12
−6L
12
−6L
6L v1 F1Y
2L2 θ1 M1
=
−6L v2 F2Y
4L2 θ 2 M2
Load and constraints (BCs) are
F2Y = − f ,
v1 = θ1 = 0
M2 = m
Reduced equation is
EI
L3
12
−6L
−6L v2 − f
=
4L2 θ 2 m
Solving this, we obtain
2
4
v2
L −2L f + 3Lm − pL /8EI
=
=
3
θ 2 6EI −3Lf + 6m − pL /6EI
(A)
These nodal values are the same as the exact solution. Note that the deflection v(x) (for
0 < x < L) in the beam by the FEM is, however, different from that by the exact solution.
The exact solution by the simple beam theory is a fourth-order polynomial of x, while
the FE solution of v is only a third-order polynomial of x.
If the equivalent moment m is ignored, we have
2
4
v2
L −2L f − pL /6EI
=
=
3
θ 2 6EI −3Lf − pL /4EI
(B)
The errors in (B) will decrease if more elements are used. The equivalent moment m is
often ignored in the FEM applications. The FE solutions still converge as more elements
are applied.
71
Beams and Frames
From the FE equation, we can calculate the reaction force and moment as
F1Y EI −12
= 3
M1 L −6L
6L v2 pL/2
=
2L2 θ 2 5 pL2/12
where the result in (A) has been used. This force vector gives the total effective nodal
forces, which include the equivalent nodal forces for the distributed lateral load p
given by
− pL/2
2
− pL /12
The correct reaction forces can be obtained as follows:
F1Y pL/2
−
=
2
M1 5 pL /12
− pL/2 pL
= 2
2
− pL /12 pL /22
Check the results: Draw the FBD for the FE model (with the equivalent nodal force
vector) and check the equilibrium condition.
pL/2
pL/2
5pL2/12
pL2/12
EXAMPLE 3.3
y
1
2
1
2
E, I
L
L
P
3 k
x
4
Given:
P = 50 kN, k = 200 kN/m, L = 3 m, E = 210 GPa, I = 2 × 10−4 m4.
Find:
Deflections, rotations, and reaction forces.
Solution
The beam has a roller (or hinge) support at node 2 and a spring support at node 3. We
use two beam elements and one spring element to solve this problem.
The spring stiffness matrix is given by
v3
k
ks =
−k
v4
−k
k
72
Finite Element Modeling and Simulation with ANSYS Workbench
Adding this stiffness matrix to the global FE equation (see Example 3.1), we have
v2
θ2
6L
−12
12
2
L
4
−
6L
24
EI
L3
Symmetry
6L
2L2
0
8L2
v1
θ1
v3
0
0
−12
−6L
12 + k ’
θ3
0
0
6L
2L2
−6L
4L2
v4
0 v1 F1Y
0 θ1 M1
0 v2 F2Y
0 θ 2 = M 2
− k ’ v3 F3Y
0 θ 3 M3
k ’ v4 F4Y
in which
k’ =
L3
k
EI
is used to simplify the notation.
We now apply the boundary conditions
v1 = θ1 = v2 = v4 = 0,
M2 = M3 = 0,
F3Y = − P
“Deleting” the first three and seventh equations (rows and columns), we have the following reduced equation:
8L2
EI
−6L
L3 2
2L
−6L
12 + k ’
−6L
2L2 θ 2 0
−6L v3 = − P
4L2 θ 3 0
Solving this equation, we obtain the deflection and rotations at nodes 2 and 3,
θ 2
3
PL2
=
−
v
3
7 L
+
(
12
7
’)
EI
k
θ
9
3
The influence of the spring k is easily seen from this result. Plugging in the given
numbers, we can calculate
θ 2 −0.002492 rad
v3 = −0.01744 m
θ −0.007475 rad
3
73
Beams and Frames
From the global FE equation, we obtain the nodal reaction forces as,
F1Y −69.78 kN
M −69.78 kN ⋅ m
1
=
F
116
.
2
kN
2
Y
F4Y 3.488 kN
Checking the results: Draw free-body diagram of the beam
50 kN
69.78 kN
1
3
2
69.78 kN · m
116.2 kN
3.488 kN
Sum the forces and moments to verify that equilibrium of the beam is satisfied.
We use the following example to show how to model frames using the general beam
elements. This example can be used to verify the FEA results when using a software
package.
EXAMPLE 3.4
Y
3000 lb
500 lb/ft
1
2
E, I, A
2
1
8 ft
3
3
4
X
12 ft
Given:
E = 30 × 10 6 psi, I = 65 in.4 , A = 6.8 in.2
Find:
Displacements and rotations of the two joints 1 and 2.
Solution
This is a problem of analyzing a frame. Members in a frame are considered to be rigidly
connected (e.g., welded together). Both forces and moments can be transmitted through
their joints. We need the general beam element (combinations of bar and simple beam elements) to model frames.
For this example, we first convert the distributed load into its equivalent nodal loads
to obtain the following FE mode.
74
Finite Element Modeling and Simulation with ANSYS Workbench
3000 lb
3000 lb
72,000 lb-in.
72,000 lb-in.
3000 lb
1
2
1
2
3
3
4
In local coordinate system, the stiffness matrix for a general 2-D beam element is
ui
EA
L
0
0
k=
EA
−
L
0
0
θi
vi
−
0
0
12EI
L3
6EI
L2
6EI
L2
4EI
L
0
0
−
12EI
L3
6EI
L2
uj
vj
EA
L
0
0
0
12EI
L3
6EI
− 2
L
−
EA
L
6EI
L2
2EI
L
−
0
0
0
12EI
L3
6EI
− 2
L
θj
6EI
L2
2EI
L
0
6EI
− 2
L
4EI
L
0
Element Connectivity Table
Element
Node i (1)
Node j (2)
1
3
4
2
1
2
1
2
3
For element 1, we have
u1
141.7
0
0
4
k1 = k′1 = 10 ×
−141.7
0
0
v1
θ1
u2
v2
0
0.784
56.4
0
−0.784
56.4
0
56.4
5417
0
−56.4
2708
−141.7
0
0
141.7
0
0
0
−0.784
−56.4
0
0.784
−56.4
θ2
0
56.44
2708
0
−56.4
5417
75
Beams and Frames
For elements 2 and 3, the stiffness matrix in local system is
u′i
212.5
0
0
4
k ′2 = k ′3 = 10 ×
−212.5
0
0
v′i
θ ′i
u ′j
v ′j
0
2.65
127
0
−2.65
127
0
127
8125
0
−127
4063
−212.5
0
0
212.5
0
0
0
−2.65
−127
0
2.65
−127
θ ′j
0
127
4063
0
−127
8125
where i = 3, j = 1 for element 2, and i = 4, j = 2 for element 3.
The transformation matrix T is
l
−m
0
T=
0
0
0
m
l
0
0
0
0
0
0
1
0
0
0
0
0
0
l
−m
0
0
0
0
m
l
0
0
0
0
0
0
1
We have l = 0, m = 1 for both elements 2 and 3. Thus,
0
−1
0
T=
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
−1
0
0
0
0
1
0
0
0
0
0
0
0
1
Using the transformation relation
k = T T k ′T
we obtain the stiffness matrices in the global coordinate system for elements 2 and 3
u3
2.65
0
−127
4
k 2 = 10 ×
−2.65
0
−127
v3
0
212.5
0
0
−212.5
0
θ3
−127
0
8125
127
0
4063
u1
v1
θ1
−2.65
0
127
2.65
0
127
0
−212.5
0
0
212.5
0
−127
0
4063
127
0
8125
76
Finite Element Modeling and Simulation with ANSYS Workbench
and
u4
2.65
0
−127
4
k 3 = 10 ×
−2.65
0
−127
v4
θ4
0
212.5
0
0
−212.5
0
−127
0
8125
127
0
4063
u2
−2.65
0
127
2.65
0
127
v2
θ2
0
−212.5
0
0
212.5
0
−127
0
4063
127
0
8125
Assembling the global FE equation and noticing the following boundary conditions
u3 = v3 = θ 3 = u4 = v4 = θ 4 = 0
F1X = 3000 lb, F2 X = 0, F1Y = F2Y = −3000 lb,
M1 = −72, 000 lb ⋅ in., M2 = 72, 000 lb ⋅ in.
we obtain the condensed FE equation
144.3
0
127
4
10 ×
−141.7
0
0
0
213.3
56.4
0
−0.784
56.4
127
56.4
13, 542
0
−56.4
2708
−141.7
0
0
144.3
0
127
0
−0.784
−56.4
0
213.3
−56.4
0 u1 3000
56.4 v1 −3000
2708 θ1 −72, 000
=
127 u2 0
−56.4 v2 −3000
13, 542 θ 2 72, 000
Solving this, we obtain
0.092 in.
u1
v −0.00104 in.
1
θ1 −0.00139 rad
=
u2 0.0901 in.
v2 −0.0018 in.
−5
θ 2 −3.88 × 10 rad
To calculate the reaction forces and moments at the two ends, we employ the element FE equations for elements 2 and 3 with known nodal displacement vectors. We
obtain
F3 X −672.7 lb
F3Y = 2210 lb
M 60, 364 lb ⋅ in.
3
77
Beams and Frames
and
F4 X −2338 lb
3825 lb
F4Y =
M 112, 641 lb ⋅ in.
4
Check the results: Draw the free-body diagram of the frame as shown below. Equilibrium
is maintained with the calculated forces and moments. Recall that the problem we
solved is the one with the equivalent loads, not the one with the distributed load. Thus,
the corresponding FBD for the FE model should be applied for verifying the results.
3000 lb
3000 lb
72,000 lb-in.
3000 lb
72,000 lb-in.
60,364 lb-in.
112,641 lb-in.
672.7 lb
2338 lb
3825 lb
2210 lb
3.6 Case Study with ANSYS Workbench
Problem Description: Steel framing systems provide cost-effective solutions for low-rise
buildings. They have high strength-to-weight ratios, and can be prefabricated and customdesigned. Consider the following two-storey building constructed with structural steel
I-beams. Determine the deformations and the stresses in the frame when a uniform load
of 50 kN/m is applied on the second floor as shown below.
1m
3m
3m
6m
6m
6m
Material: Structural steel
Line pressure: 50 kN/m
I-beam size: W356 × 171
Beam depth = 355.6 mm
Flange width = 171.5 mm
Web thickness = 11.5 mm
Flange thickness = 7.3 mm
Solution
To solve the problem with ANSYS® Workbench, we employ the following steps:
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Steelframe.wbpj.”
78
Finite Element Modeling and Simulation with ANSYS Workbench
Step 2: Create a Static Structural (ANSYS) Analysis System
Drag the Static Structural (ANSYS) icon from the Analysis Systems Toolbox window
and drop it inside the highlighted green rectangle in the Project Schematic window to create a standalone static structural analysis system.
Step 3: Launch the DesignModeler Program
Double-click the Geometry cell to launch DesignModeler, and select “Meter” as
length unit in the Units pop-up window. Ensure Line Bodies is selected in the
Properties of Schematic A3: Geometry window.
Turn on ‘Line
Bodies’ to
enable beam
modeling.
Step 4: Create Line Sketch
Click the Sketching tab and select Settings. Turn on Show in 2D and Snap under
Grid options. Use the default value of “5 m” for Major Grid Spacing and “5” for
Minor-Steps per Major.
79
Beams and Frames
Click a start point and then an end point in the Graphics window to draw a line.
Draw 10 lines as shown in the sketch below. After completion, click Generate to
create a line sketch.
Click the first point and
then click the second
point to draw a line.
2
1
7
9
4
3
8
10
6
5
Step 5: Create Line Body from Sketch
Check off the Grid options under Settings of Sketching Toolboxes. Switch to the
Modeling tab. Note that a new item named Sketch1 now appears underneath
XYPlane in the Tree Outline.
Select Lines from Sketches from the Concept drop-down menu. Click Zoom to Fit.
Zoom to Fit
80
Finite Element Modeling and Simulation with ANSYS Workbench
Select Sketch1 from the Tree Outline and click Apply to confirm on the Base Objects
selection in the Details of Line1. Click Generate to complete the line body
creation.
Step 6: Create Line Body through Translation
Select Body Operation from the Create drop-down menu. A new item named
BodyOp1 is now added to the Tree Outline.
81
Beams and Frames
In the Details of BodyOp1, click anywhere on the Type cell and select Translate from
the drop-down menu.
Select the line body from the Graphics window and then click Apply to confirm
on the Bodies selection in the Details of BodyOp1. After completion, change the
Preserve Bodies? selection to Yes. This will help preserve a copy of the selected
line body at the current location while translating it to a new location.
Click
anywhere on
an edge to
select the line
body.
82
Finite Element Modeling and Simulation with ANSYS Workbench
In the Details of BodyOp1, change the Direction Definition to Coordinates, and enter
“−6” for the Z Offset. Click Generate. After completion, the line body will be
copied backward by 6 m.
Step 7: Create Line Body from Points
Select Lines From Points from the Concept drop-down menu. After completion, a
new item named Line2 is added to the Tree Outline.
83
Beams and Frames
In the Graphics window, select a start point and Ctrl-select an end point to draw
a line. Draw six new lines connecting the two planar frames as shown below.
Click Apply to confirm on the Point Segments selection in the Details of Line2.
Click Generate to complete the line creation.
Click on the
first point
and Ctrlclick on the
second
point to
1
4
2
3
5
6
Note that the line bodies created in all previous steps now merge into a single line
body. Check off the Cross Section Alignments from the View drop-down menu
to switch-off the display of local coordinate systems.
84
Finite Element Modeling and Simulation with ANSYS Workbench
Step 8: Create a Cross Section
Select a Cross Section of I Section from the Concept drop-down menu. A new item
named I1 is now added underneath the Cross Section in the Tree Outline.
In the Details of I1 under Dimensions, enter “0.1715” for W1 and W2, “0.3556” for
W3, “0.0073” for t1 and t2, and “0.0115” for t3.
Beams and Frames
Step 9: Assign Cross Section to Line Body
Select the Line Body underneath 1Part, 1 Body in the Tree Outline. In the Details of
Line Body, assign I1 to the Cross Section selection.
Turn on the Cross Section Solids from the View drop-down menu to view the frame
as a solid structure.
85
86
Finite Element Modeling and Simulation with ANSYS Workbench
Note that some I-beams in the above structure are used as H-beams. To fix the misaligned cross sections, turn on the Edge Selection Filter and select the eight line
edges shown below from the Graphics window. In the Details of Line-Body Edges,
enter “90” for Rotate to turn the beams 90° about their neutral axes.
Click to turn on/off the Edge Selection Filter.
2
1
5
Ctrl-click lines
to make
multiple
selections.
4
6
3
8
7
The adjusted frame shown below now has all the I-beams oriented in the strongaxis configuration. This completes the geometry creation of a frame structure.
Click Close DesignModeler to exit the program.
Beams and Frames
Step 10: Launch the Static Structural (ANSYS) Program
Double-click the Model cell to launch the Static Structural (ANSYS) program. From
its Units drop-down menu, select Metric (m, Kg, N, s, V, A). Note that in the
Details of “Line Body” the material is assigned to Structural Steel by default.
Step 11: Generate Mesh
In the Details of “Mesh,” enter “0.2 m” for the Element Size. In the Outline of Project,
right-click on Mesh and select Generate Mesh.
87
88
Finite Element Modeling and Simulation with ANSYS Workbench
After completion, the meshed structure appears in the Graphics window. You may
deselect the Ruler from the View drop-down menu to turn off the ruler display
in the Graphics window.
Step 12: Apply Boundary Conditions
In the Outline of Project, right-click on Static Structural (A5) and select Insert and
then Fixed Support. After completion, a Fixed Support item is added underneath
Static Structural (A5) in the project outline tree.
89
Beams and Frames
Select the six points as shown below in the Graphics window. In the Details of “Fixed
Support,” click Apply to confirm on the Geometry selection. After completion,
a Fixed Support boundary condition will be added to the selected six points.
Turn on Vertex filter.
Ctrl-click
points to
make
multiple
selections.
2
4
1
6
3
5
Step 13: Apply Loads
In the Outline of Project, right-click on Static Structural (A5) and select Insert and
then Line Pressure.
90
Finite Element Modeling and Simulation with ANSYS Workbench
Select the line as shown below in the Graphics window. In the Details of “Line
Pressure,” click Apply to confirm on the Geometry selection.
Click a line to
make a
selection.
In the Details of “Line Pressure,” change the Define By selection to Components and
enter “-50000” for the Y Component. A downward red arrow will appear on the
selected line in the Graphics window.
91
Beams and Frames
Repeat the steps of adding a line pressure, and insert the second Line Pressure item
underneath Static Structural (A5) in the Project Outline tree. Apply the same
exact load to the selected line highlighted in the following figure.
Click on the line
and apply it to
the geometry
selection of line
Repeat the steps until the same load is applied to all seven edges highlighted in
the figure below.
6
1
5
4
7
2
3
Alternative Procedure: Note that the Line Pressure item in the Outline can be copied and pasted under Static Structural (A5) for repeated use. To make a copy,
right-click on Line Pressure and select Copy from the menu. To paste, right-click
on the Static Structural (A5) and select Paste. Remember to apply each newly
pasted Line Pressure to a different line edge on the Geometry selection in the
Details of “Line Pressure” until the same load is applied to all seven edges.
92
Finite Element Modeling and Simulation with ANSYS Workbench
Step 14: Retrieve Solution
Insert a Total Deformation item by right-clicking on Solution (A6) in the Outline tree.
Insert a Beam Tool item by right-clicking on Solution (A6) in the Outline tree.
Beams and Frames
Right-click on Solution (A6) in the Outline tree and select Solve. The program will
start to solve the model.
After completion, click Total Deformation in the Outline to review the total deformation results.
93
94
Finite Element Modeling and Simulation with ANSYS Workbench
Click Direct Stress under Beam Tool in the Outline to review the axial stress results
in beams.
Click Minimum Combined Stress under Beam Tool to retrieve the linear combination
of the Direct Stress and the Minimum Bending Stress results in beams.
95
Beams and Frames
Click Maximum Combined Stress under Beam Tool to retrieve the linear combination
of the Direct Stress and the Maximum Bending Stress results in beams.
Close the Static Structural (ANSYS) program. Save project and exit Workbench.
Modeling tips: A group of points can be constructed in DesignModeler through a text
file that contains the following five columns: group number, point number, x-coordinate,
y-coordinate, and z-coordinate. The data columns can be separated by spaces or Tabs.
For example, to create five points with their coordinates given as x(i), y(i), and z(i), type
the data in the following form in Notepad or WordPad, and save as a text file, for example,
“points.txt,” in local disk:
1
1
1
1
1
1
2
3
4
5
x(1)
x(2)
x(3)
x(4)
x(5)
y(1)
y(2)
y(3)
y(4)
y(5)
z(1)
z(2)
z(3)
z(4)
z(5)
Then, follow the steps described in the following figure:
Select the text file that contains the
point coordinate data.
Click point under the create drop-down
menu of design modeler.
Click
generate to
create the
points in the
graphics
window.
96
Finite Element Modeling and Simulation with ANSYS Workbench
3.7 Summary
This chapter is devoted to the beam element and frame analysis. We have studied the
beam element, which can be used in frame analysis. The concept of the shape functions is
further explored and the derivations of the stiffness matrix using the energy approach are
emphasized. Treatment of distributed loads is discussed and several examples are studied. A two-story building structure with I-beams is analyzed using ANSYS Workbench.
It provides useful modeling techniques in constructing concept line models, and shows
step-by-step how Workbench can be used to determine the deformation and stresses in
beams and frames.
3.8 Review of Learning Objectives
Now that you have finished this chapter, you should be able to
1.
2.
3.
4.
5.
Set up simplified finite element models for beams and frames.
Derive the element stiffness matrix for plane beams using direct/energy approach.
Explain the concept of shape functions and their characteristics for beam elements.
Find the equivalent nodal forces of distributed loads on beams.
Determine the deflection and rotation at a point of a beam using hand calculation
to verify the finite element solutions.
6. Apply the general beam element stiffness matrix to the analysis of simple frames.
7. Create line models from concept points, sketches, or by body translation in
Workbench.
8. Perform static structural analyses on beams and frames using Workbench.
PROBLEMS
3.1 Using Equation 3.11, derive the results of the equivalent nodal forces and
moments for a beam element with uniformly distributed lateral load.
3.2 The cantilever beam is supported by a spring at the end as shown in the figure.
Using FEM, determine the deflection and rotation at the node 2.
y
P
E, I
1
2
L
k
3
x
97
Beams and Frames
3.3 Using FEM, determine the nodal displacement, rotations, and reaction forces
for the propped cantilever beam shown below. The beam is assumed to have
constant EI and length 2L. It is supported by a roller at mid length and is built
in at the right end.
Y
P
1
1
2
E, I
2
3
L
X
L
3.4 The 2-D frame is supported as shown in the figure. Constants E, A, I of the beam
and the length L are given. Using FEM, determine the displacement and rotation at node 2.
Y
M
1
2
L
X
L
3
3.5 The plane frame is subjected to the uniformly distributed load and is fixed
at the ends as shown in the figure. Assume E = 30 × 106 psi, A = 100 in.2, and
I = 1000 in.4 for both elements of the frame. Using FEM, find the displacement
and rotation of node 2.
1000 lb/ft
2
Y
3
40 ft
1
45°
X
30 ft
3.6 Using ANSYS Workbench, solve the frame problem in Problem 3.5, and determine the reaction forces and moments at the two fixed ends.
98
Finite Element Modeling and Simulation with ANSYS Workbench
3.7 The physical construction of two representative beam connections is shown in
the figure below, where an I-beam is connected to a floor slab through a slotted
bolt hole in (a), and is connected to a column through an angle bracket in (b).
What simplified support conditions would you use to represent the physical
construction in the two cases and explain your reasons?
(a)
(b)
3.8 The bike frame shown in the figure below has hollow circular tubes (24 mm
outer diameter and 2 mm thick) and is made of aluminum alloy. Use ANSYS
Workbench to determine the deformation and stresses of the frame members.
Point coordinate (mm)
x
y
z
1
400 400
0
2 –150 350
0
3
450 300
0
4
0
0
0
5
600
50 50
600
6
50 –50
7 –450
50 50
8 –450
50 –50
80 N
540 N
1
2
8
3
6
4
7
5
180 N
3.9 The hoist frame shown in the following figure has 76 mm wide and 3 mm thick
square tubes and is made of structural steel. Use ANSYS Workbench to determine the deformation and the stresses in the frame.
0.5 m
1
4
0.5 m
2
5 6
8000 N
3
7
2m
8 9
0.2 m 0.2 m
10
14
11
15
12
16
13
17
1m
1.2 m
1.6 m
0.3 m
99
Beams and Frames
3.10 The frame structure is constructed with structural steel I-beams of given crosssection dimensions and is used to support a uniform load of 6 kN/m as shown
below. Use ANSYS Workbench to determine the deformation and stresses in the
frame members.
6 kN/m
25 mm
0.3 m
6 kN/m
0.3 m
5 mm
80 mm
5 mm
0.5 m
1.2 m
3.11 The following bridge structure is used to carry a uniform load of 50 kN/m on
the bridge surface. Use ANSYS Workbench to determine the deformation and
stresses in the bridge if structural steel I-beams of the given cross-section are
used.
50 kN/m
1m
4.06 in.
0.375 in.
0.2 m
0.28 in.
2m
0.2 m 1 m
1.5 m
1.5 m
1m
0.2 m
4.16 in.
4
Two-Dimensional Elasticity
4.1 Introduction
Many structures that are three-dimensional can be satisfactorily treated as two-dimensional problems (Figure 4.1). This chapter introduces you to the use of the finite element
method for deformation and stress analyses of two-dimensional elasticity problems. First,
the basic equations in plane elasticity theory are reviewed. Then several types of 2-D finite
elements for plane elasticity analysis are presented. Applications of these elements are
demonstrated and their accuracies and efficiencies are discussed. This presentation is
followed by an example illustrating analysis of a 2-D elasticity problem using ANSYS®
Workbench.
4.2 Review of 2-D Elasticity Theory
In general, the stresses and strains at any point in a structure consist of six independent
components, that is (Figure 4.2),
σ x , σ y , σ z , τ xy , τ yz , τ zx
for stresses, and
ε x , ε y , ε z , γ xy , γ yz , γ zx
for strains.
Under certain conditions, the state of stresses and strains can be simplified. A 3-D stress
analysis can, therefore, be reduced to a 2-D analysis. There are two general types of models
involved in this 2-D analysis: plane stress and plane strain.
4.2.1 Plane Stress
In the plane stress case, any stress component related to the z direction is zero, that is,
σ z = τ yz = τ zx = 0
( ε z ≠ 0)
(4.1)
101
102
Finite Element Modeling and Simulation with ANSYS Workbench
(b)
(a)
River
A rotating disk
A reservoir dam
FIGURE 4.1
Examples of 2-D elasticity problems.
σy
τxy
τyz
τzx
y
σx
σz
x
z
FIGURE 4.2
Stress components at a point in a structure.
A thin planar structure with constant thickness and loading within the plane of the
structure (xy-plane) can be regarded as a plane stress case (Figure 4.3).
4.2.2 Plane Strain
In the plane strain case, any strain component related to the z direction is zero, that is,
ε z = γ yz = γ zx = 0
(σ z ≠ 0 )
(4.2)
A long structure with a uniform cross section and transverse loading along its thickness
(z-direction), such as a tunnel or a dam, can be regarded a plane strain case (Figure 4.4).
As illustrated in the above figures, structures in plane stress or plane strain conditions can be modeled as 2-D boundary value problems. At an arbitrary point, the inplane ­displacement field may be defined by two components (u and v), the in-plane strain
field by three components (ε x , ε y , γ xy ), and the in-plane stress field by three components
(σ x , σ y , τ xy ). The following is a review of equations describing how the field variables are
related.
103
Two-Dimensional Elasticity
(a)
(b)
z
y
y
x
x
FIGURE 4.3
Plane stress condition: (a) structure with in-plane loading; and (b) 2-D model.
(a)
y
(b)
x
z
y
x
FIGURE 4.4
Plane strain condition: (a) structure with uniform transverse load; and (b) 2-D model.
4.2.3 Stress–Strain (Constitutive) Equations
For linear elastic isotropic materials, we have the following strain–stress relation for the
plane stress case:
ε x 1/E
ε y = −ν/E
γ 0
xy
−ν/E
1/E
0
0 σx
0 σy +
1/G τ xy
ε x0
εy0
γ
xy 0
(4.3)
or,
ε = E −1σ + ε 0
where ε0 is the initial strain (e.g., due to a temperature change), E Young’s modulus, ν
Poisson’s ratio, and G the shear modulus. Note that
G=
for homogeneous and isotropic materials.
E
2(1 + ν)
(4.4)
104
Finite Element Modeling and Simulation with ANSYS Workbench
We can also express stresses in terms of strains by solving the above equation
σx
1
E
ν
σy =
2
τ 1 − ν 0
xy
εx
εy −
(1 − ν)/ 2 γ xy
ν
1
0
0
0
ε x 0
ε y 0
γ
xy 0
(4.5)
or,
σ = Eε + σ 0
where σ 0 = − Eε 0 is the initial stress.
For plane strain case, we need to replace the material constants in the above equations in
the following fashion:
E→
E
;
1 − ν2
ν→
ν
;
1− ν
G→G
(4.6)
For example, the stress is related to strain by
σx
1 − ν
E
σ
=
y
ν
(
1
)(
1
2
)
+
ν
−
ν
τ
0
xy
ν
1− ν
0
0
εx
0
εy −
(1 − 2 ν)/2 γ xy
ε x 0
ε y 0
γ
xy 0
in the plane strain case.
Initial strain due to a temperature change (thermal loading) is given by the following for
the plane stress case
ε x 0 α ∆T
ε y 0 = α ∆T
γ 0
xy 0
(4.7)
where α is the coefficient of thermal expansion, ΔT the change of temperature. For the
plane strain case, α should be replaced by (1 + ν)α in Equation 4.7. Note that if the structure is free to deform under thermal loading, there will be no (elastic) stresses in the
structure.
4.2.4 Strain and Displacement Relations
For small strains and small rotations, we have,
εx =
∂u
,
∂x
εy =
∂v
,
∂y
γ xy =
∂u
∂v
+
∂y
∂x
105
Two-Dimensional Elasticity
In matrix form, we write
ε x ∂/∂ x
εy = 0
γ ∂/∂ y
xy
0
u
∂/∂ y , or ε = Du
v
∂/∂ x
(4.8)
From this relation, we know that, if the displacements are represented by polynomials,
the strains (and thus stresses) will be polynomials of an order that is one order lower than
the displacements.
4.2.5 Equilibrium Equations
In plane elasticity, the stresses in the structure must satisfy the following equilibrium
equations:
∂σ x ∂τ xy
+ fx = 0
+
∂x
∂y
∂τ xy ∂σ y
+
+ fy = 0
∂x
∂y
(4.9)
where fx and f y are body forces (forces per unit volume, such as gravity forces). In the FEM,
these equilibrium conditions are satisfied in an approximate sense.
4.2.6 Boundary Conditions
The boundary S of the 2-D region can be divided into two parts, Su and St (Figure 4.5). The
boundary conditions (BCs) can be described as
u = u,
tx = tx ,
v = v,
on Su
ty = ty ,
on St
(4.10)
in which tx and ty are tractions (stresses on the boundary) and the barred quantities are
those with known values.
4.2.7 Exact Elasticity Solution
The exact solution (displacements, strains, and stresses) of a given problem must satisfy
the constitutive relations, equilibrium equations, and compatibility conditions (structures
ty
p
tx
y
St
x
FIGURE 4.5
Boundary conditions for a structure.
Su
106
Finite Element Modeling and Simulation with ANSYS Workbench
should deform in a continuous manner, with no cracks or overlaps in the obtained displacement fields).
EXAMPLE 4.1
y
p
x
A plate is supported and loaded with distributed force p as shown in the figure. The
material constants E and ν are known.
The exact solution for this simple problem is found to be as follows:
Displacement:
u=
p
p
x , v = −ν y
E
E
Strain:
εx =
p
p
, ε y = −ν , γ xy = 0
E
E
Stress:
σ x = p, σ y = 0, τ xy = 0
Exact (or analytical) solutions for simple problems are numbered (suppose there is a
hole in the plate or the roller supports are replaced by fixed ones). That is why we need
the FEM for solutions of 2-D elasticity problems in general.
4.3 Modeling of 2-D Elasticity Problems
For a plane stress or plane strain analysis, we model only a 2-D region, that is, a planar
surface or cross section of the original 3-D structure. The region then needs to be divided
into an element discretization made of triangles, quadrilaterals, or a mixture of both. The
element behaviors need to be specified to set up the problem type as either plane stress
or plane strain. The discretization can be structured (mapped mesh on a three-sided or
four-sided surface region with equal numbers of element divisions for the opposite sides)
or unstructured (free mesh), as shown in Figure 4.6. A structured surface mesh has r­ egular
107
Two-Dimensional Elasticity
(a)
(b)
FIGURE 4.6
Finite element discretization: (a) unstructured (free) mesh; and (b) structured (mapped) mesh.
(a)
(b)
(c)
FIGURE 4.7
Support conditions in 2-D: (a) and (b) ineffective constraints; (c) effective constraints.
connectivity that can be described as a 2-D array, and it generally has better solution reliability compared to a free mesh.
Mesh properties such as the mesh density and the element shape quality are important
factors that affect the solution accuracy and efficiency. Finer mesh is usually needed in
areas where a high stress/strain is expected. The type of finite element employed also
plays a vital role in the analysis, as will be discussed in the next section.
After meshing is complete, the surface discretization is represented in terms of nodal
coordinates as well as element connectivity showing how different elements are connected
together through nodes. All types of loads (distributed forces, concentrated forces and
moments, and so on) are converted into point forces acting at individual nodes. Boundary
conditions are no exception, and will be converted into nodal DOF constraints. Remember
that support conditions need to suppress rigid-body motion, and the minimum conditions
in 2-D include restraining the body with displacement in x, y directions and rotation about
the z-axis. Figure 4.7 gives examples of effective and ineffective support conditions in 2-D.
Explain why a case is effective or not.
4.4 Formulation of the Plane Stress/Strain Element
In this section, we formulate the element equations based on the plane elasticity theory. A
general formula for the stiffness matrix is derived first, followed by presentation of several
common element types for 2-D elasticity problems.
108
Finite Element Modeling and Simulation with ANSYS Workbench
4.4.1 A General Formula for the Stiffness Matrix
Displacements (u, v) in a plane element can be interpolated from nodal displacements
(ui, vi) using shape functions Ni as follows:
u N1
=
v 0
0
N1
u1
v
1
u
2 or
v
2
0
N2
N2
0
u = Nd
(4.11)
where N is the shape function matrix, u the displacement vector, and d the nodal displacement vector. Here, we have assumed that u depends on the nodal values of u only, and v
on nodal values of v only.
From strain–displacement relation (Equation 4.8), the strain vector is
ε = Du = DNd, or ε = Bd
(4.12)
where B = DN is the strain–displacement matrix.
Consider the strain energy stored in an element,
U=
1
1
σ T ε dV =
2
2
∫
V
=
1
2
∫ (σ ε
x
x
)
+ σ y ε y + τ xy γ xy dV
V
1
1
∫ (Eε) ε dV = 2 ∫ ε Eε dV = 2 d ∫ B EB dVd
T
V
T
V
T
T
V
1
= dT kd
2
From this, we obtain the general formula for the element stiffness matrix
k=
∫ B EB dV
T
(4.13)
V
Note that unlike the 1-D cases, E here is a matrix which is given by the stress–strain relation (e.g., Equation 4.5 for plane stress).
The stiffness matrix k defined by Equation 4.13 is symmetric since E is symmetric. Also
note that given the material property, the behavior of k depends on the B matrix only, which
in turn depends on the shape functions. Thus, the quality of finite elements in representing
the behavior of a structure is mainly determined by the choice of shape functions. Most
commonly employed 2-D elements are linear or quadratic triangles and quadrilaterals.
4.4.2 Constant Strain Triangle (CST or T3)
This is the simplest 2-D element (Figure 4.8), which is also called linear triangular element.
For this element, we have three nodes at the vertices of the triangle, which are numbered
around the element in the counterclockwise direction. Each node has two DOFs (can move
109
Two-Dimensional Elasticity
v3
3
(x3, y3)
u3
v
y
v1
(x, y)
1
(x1, y1)
v2
u
2
(x2, y2)
u1
u2
x
FIGURE 4.8
Linear triangular element (T3).
in the x and y directions). The displacements u and v are assumed to be linear functions
within the element, that is
u = b1 + b2 x + b3 y , v = b4 + b5 x + b6 y
(4.14)
where bi (i = 1, 2, …, 6) are constants. From these, the strains are found to be
ε x = b2 , ε y = b6 , γ xy = b3 + b5
(4.15)
which are constant throughout the element. Thus, we have the name “constant strain triangle” (CST).
Displacements given by Equation 4.14 should satisfy the following six equations:
u1 = b1 + b2 x1 + b3 y1
u2 = b1 + b2 x2 + b3 y 2
v3 = b4 + b5 x3 + b6 y 3
Solving these equations, we can find the coefficients b1, b2, …, and b6 in terms of nodal
displacements and coordinates. Substituting these coefficients into Equation 4.14 and rearranging the terms, we obtain
u N1
=
v 0
0
N1
N2
0
0
N2
N3
0
u1
v
1
0 u2
N 3 v2
u3
v3
(4.16)
110
Finite Element Modeling and Simulation with ANSYS Workbench
where the shape functions (linear functions in x and y) are
1
{(x2 y3 − x3 y2 ) + ( y2 − y3 )x + (x3 − x2 )y}
2A
1
N2 =
{(x3 y1 − x1y3 ) + ( y3 − y1 )x + (x1 − x3 )y}
2A
1
N3 =
{(x1y2 − x2 y1 ) + ( y1 − y2 )x + (x2 − x1 )y}
2A
N1 =
(4.17)
and
A=
1
1
det 1
2
1
x1
x2
x3
y1
y2
y 3
(4.18)
is the area of the triangle.
Using the strain–displacement relation (4.8), results (4.16) and (4.17), we have
εx
y 23
1
0
ε y = Bd =
2A
γ
x
32
xy
0
x32
y 23
y 31
0
x13
0
x13
y 31
y12
0
x21
u1
v
1
0
u2
x21
v2
y12
u3
v3
(4.19)
where xij = xi − xj and yij = yi − yj (i, j = 1, 2, 3). Again, we see constant strains within the element. From stress–strain relation (e.g., Equation 4.5), we see that stresses obtained using
the CST element are also constant.
Applying the formula in Equation 4.13, we obtain the element stiffness matrix for the
CST element
k=
∫ B EB dV = tA(B EB)
T
T
(4.20)
V
in which t is the thickness of the element. Notice that k for CST is a 6 × 6 symmetric matrix.
Both the expressions of the shape functions in Equation 4.17 and their derivations are
lengthy and offer little insight into the behavior of the element.
We introduce the natural coordinates (ξ, η) on the triangle where ξ = A1/A, η = A2/A with
A1 and A2 being the areas of the smaller triangles indicated in Figure 4.9. Then the shape
functions can be represented simply by
N1 = ξ , N 2 = η,
N3 = 1 − ξ − η
(4.21)
Notice that,
N1 + N 2 + N 3 = 1
(4.22)
111
Two-Dimensional Elasticity
ξ=0
η=0
3
η=b
η=1
ξ=a
A2
ξ=1
A1
(a, b)
2
1
FIGURE 4.9
The natural coordinate system defined on the triangle.
which ensures that the rigid-body translation is represented by the chosen shape functions. Also, as in the 1-D case,
1,
Ni =
0,
at node i ;
at other nodes
(4.23)
and varies linearly within the element. The plot for shape function N1 is shown in Figure
4.10. N2 and N3 have similar features.
We have two coordinate systems for the element: the global coordinates (x, y) and the
natural (local) coordinates (ξ, η). The relation between the two is given by
x = N1x1 + N 2 x2 + N 3 x3
y = N 1 y1 + N 2 y 2 + N 3 y 3
(4.24)
or,
x = x13 ξ + x23 η + x3
(4.25)
y = y13 ξ + y 23 η + y 3
where xij = xi − xj and yij = yi − yj (i, j = 1, 2, 3) as defined earlier.
ξ=0
3
N1
ξ=1
1
1
FIGURE 4.10
Plot of the shape function N1 for T3 element.
2
112
Finite Element Modeling and Simulation with ANSYS Workbench
Displacement u or v on the element can be viewed as functions of (x, y) or (ξ, η). Using the
chain rule for derivatives, we have
∂u ∂x
∂ξ ∂ξ
=
∂u ∂x
∂ η ∂ η
∂ y ∂u
∂u
∂x
∂ ξ ∂ x
= J
∂ y ∂u
∂u
∂ y
∂ η ∂ y
(4.26)
where J is called the Jacobian matrix of the transformation.
From Equation 4.25, we calculate
y13
,
y 23
x13
J=
x23
J−1 =
1 y 23
2 A − x23
− y13
x13
(4.27)
where det J = x13 y 23 − x23 y13 = 2 A has been used (A is the area of the triangle).
From Equations 4.26, 4.27, 4.16, and 4.21, we have
∂u
∂x
1 y 23
=
u
∂
2
A − x23
∂ y
∂u
− y13 ∂ ξ
1 y 23
=
x13 ∂ u 2 A − x23
∂ η
− y13 u1 − u3
x13 u2 − u3
(4.28)
Similarly,
∂v
∂x
1 y 23
=
v
∂
2
A − x23
∂ y
− y13 v1 − v3
x13 v2 − v3
(4.29)
Using the results in Equations 4.28 and 4.29, and the relations ε = Du = DNd = Bd, we
obtain the strain–displacement matrix,
y 23
1
B=
0
2A
x32
0
x32
y 23
y 31
0
x13
0
x13
y 31
y12
0
x21
0
x21
y12
(4.30)
which is the same as we derived earlier in Equation 4.19.
We should note the following about the CST element:
• Use in areas where the strain gradient is small.
• Use in mesh transition areas (fine mesh to coarse mesh).
• Avoid using CST in stress concentration or other crucial areas in the structure,
such as edges of holes and corners.
• Recommended only for quick and preliminary FE analysis of 2-D problems.
113
Two-Dimensional Elasticity
4.4.3 Quadratic Triangular Element (LST or T6)
This type of element (Figure 4.11) is also called quadratic triangular element. There are six
nodes on this element: three corner nodes and three midside nodes. Each node has two
DOFs as before. The displacements (u, v) are assumed to be quadratic functions of (x, y),
u = b1 + b2 x + b3 y + b4 x 2 + b5 xy + b6 y 2
v = b7 + b8 x + b9 y + b10 x 2 + b11xy + b12 y 2
(4.31)
where bi (i = 1, 2, …, 12) are constants. From these, the strains are found to be
ε x = b2 + 2b4 x + b5 y
ε y = b9 + b11x + 2b12 y
(4.32)
γ xy = (b3 + b8 ) + (b5 + 2b10 )x + (2b6 + b11 )y
which are linear functions. Thus, we have the “linear strain triangle” (LST), which provides better results than the CST.
In the natural coordinate system, the six shape functions for the LST element are
N1 = ξ(2ξ − 1)
N 2 = η(2η − 1)
N 3 = ζ(2ζ − 1)
(4.33)
N4 = 4 ξ η
N 5 = 4 ηζ
N6 = 4 ζ ξ
in which ζ = 1 − ξ − η. Each of these six shape functions represents a quadratic form on the
element as shown in Figure 4.12.
v3
3
u3
v6
u6
5
6
u5
v2
v1
y
1
u1
x
FIGURE 4.11
Quadratic triangular element (T6).
v5
4
v4
u4
2
u2
114
Finite Element Modeling and Simulation with ANSYS Workbench
ξ=0
ξ = 1/2
ξ=1
3
5
6
N1
1
2
4
1
FIGURE 4.12
Plot of the shape function N1 for T6 element.
Displacements can be written as
6
u=
6
∑N u ,
v=
i i
i =1
∑N v
(4.34)
i i
i =1
The element stiffness matrix is still given by k = ∫ V BT EB dV, but here BTEB is quadratic
in x and y. In general, the integral has to be computed numerically.
4.4.4 Linear Quadrilateral Element (Q4)
There are four nodes at the corners of the quadrilateral element (Figure 4.13). In the natural
coordinate system (ξ, η), the four shape functions are
1
(1 − ξ)(1 − η)
4
1
N 2 = (1 + ξ)(1 − η)
4
1
N 3 = (1 + ξ)(1 + η)
4
1
N 4 = (1 − ξ)(1 + η)
4
N1 =
η
v4
η=1
4
y
η = –1
x
FIGURE 4.13
Linear quadrilateral element (Q4).
v3
u3
3
u4
v1
1
(4.35)
2
v2
u2
u1
ξ = –1
ξ
ξ=1
115
Two-Dimensional Elasticity
Note that ∑ i4=1 N i = 1 at any point inside the element, as expected.
The displacement field is given by
4
u=
∑
4
v=
N iui ,
i =1
∑N v
(4.36)
i i
i =1
which are bilinear functions over the element. The stress and strain fields are constant on
this type of elements.
4.4.5 Quadratic Quadrilateral Element (Q8)
This is the most widely used element for 2-D problems due to its high accuracy in analysis
and flexibility in modeling.
There are eight nodes for this element (Figure 4.14), four corners nodes and four midside
nodes. In the natural coordinate system (ξ, η), the eight shape functions are
N1 =
N2 =
N3 =
N4 =
N5 =
N6 =
N7 =
N8 =
1
(1 − ξ)( η − 1)(ξ + η + 1)
4
1
(1 + ξ)( η − 1)( η − ξ + 1)
4
1
(1 + ξ)(1 + η)(ξ + η − 1)
4
1
(ξ − 1)( η + 1)(ξ − η + 1)
4
1
(1 − η)(1 − ξ 2 )
2
1
(1 + ξ)(1 − η2 )
2
1
(1 + η)(1 − ξ 2 )
2
1
(1 − ξ)(1 − η2 )
2
(4.37)
η
η=1
5
1
η = –1
x
FIGURE 4.14
Quadratic quadrilateral element (Q8).
ξ
6
8
y
3
7
4
ξ = –1
2
ξ=1
116
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 4.15
Analysis of composite materials (mesh and contour stress plots).
Again, we have ∑ i8=1 N i = 1 at any point inside the element.
The displacement field is given by
8
u=
∑
8
v=
N iui ,
i =1
∑N v
i i
(4.38)
i =1
which are quadratic functions over the element. Strains and stresses over a quadratic
quadrilateral element are linear functions, which are better representations. A model of
fiber-reinforced composite materials using the Q8 elements is shown in Figure 4.15.
We need to note the following when applying the 2-D elements:
• Q4 and T3 are usually used together in a mesh with linear elements.
• Q8 and T6 are usually applied in a mesh composed of quadratic elements.
• Linear elements are good for deformation analysis, that is, when global responses
need to be determined.
• Quadratic elements are preferred for stress analysis, because of their high accuracy and the flexibility in modeling complex geometry, such as curved boundaries.
4.4.6 Transformation of Loads
Concentrated load (point forces), surface traction (pressure loads), and body force (weight)
are the main types of loads applied to a structure. Both traction and body forces need to be
converted into nodal forces in the FE model, since they cannot be applied to the FE model
directly. The conversions of these loads are based on the same idea (the equivalent-work
concept), which we have used for the cases of bar and beam elements.
Suppose, for example, we have a linearly varying traction q on a Q4 element edge, as
shown in the Figure 4.16. The traction is normal to the boundary. Using the local (tangential) coordinate s, we can write the work done by the traction q as
L
Wq =
1
t un (s)q(s)ds
2
∫
0
117
Two-Dimensional Elasticity
qB
q
qA
fB
fA
s
B
L
A
B
A
FIGURE 4.16
Traction applied on the edge of a Q4 element.
where t is the thickness, L the side length, and un the component of displacement normal
to the edge AB.
For the Q4 element (linear displacement field), we have
un (s) = (1 − s/L)unA + (s/L)unB
The traction q(s), which is also linear, is given in a similar way
q(s) = (1 − s/L)qA + (s/L)qB
Thus, we have
L
1
Wq = t unA
2
0
∫
1 − s/L
unB
1 − s/L
s/L
L
=
1
unA
2
(1 − s/L)2
unB t
(s/L)(1 − s/L)
0
=
1
unA
2
unB
=
1
unA
2
fA
unB
fB
∫
tL 2
6 1
qA
s/L ds
qB
(s/L)(1 − s/L) qA
ds
(s/L)2
qB
1 qA
2 qB
and hence the equivalent nodal force vector is
f A tL 2
=
6 1
fB
1 qA
2 qB
Note, for constant q, we have
f A qtL 1
=
2 1
fB
118
Finite Element Modeling and Simulation with ANSYS Workbench
For quadratic elements (either triangular or quadrilateral), the traction is converted into
forces at three nodes along the edge, instead of two nodes. Traction tangent to the boundary, as well as body forces, are converted into nodal forces in a similar way.
4.4.7 Stress Calculation
The stress in an element is determined by the following relation:
σx
εx
σ y = E ε y = EBd
τ
γ
xy
xy
(4.39)
where B is the strain-nodal displacement matrix and d is the nodal displacement vector,
which is known for each element once the global FE equation has been solved.
Stresses can be evaluated at any point inside the element (such as the center) or at the
nodes. Contour plots are usually used in FEA software packages (during postprocess) for
users to visually inspect the stress results.
4.4.7.1 The von Mises Stress
The von Mises stress is the effective or equivalent stress for 2-D and 3-D stress analysis. For
a ductile material, the stress level is considered to be safe, if
σ e ≤ σY
where σe is the von Mises stress and σY the yield stress of the material. This is a generalization of the 1-D (experimental) result to 2-D and 3-D situations.
The von Mises stress is defined by
σe =
1
( σ 1 − σ 2 )2 + ( σ 2 − σ 3 ) 2 + ( σ 3 − σ 1 ) 2
2
(4.40)
in which σ1, σ2 and σ3 are the three principle stresses at the considered point in a structure.
For 2-D problems, the two principle stresses in the plane are determined by
2
σ 1P =
σx + σy
σx − σy
2
+
+ τ xy
2
2
2
σ 2P
σx + σy
σx − σy
2
=
−
+ τ xy
2
2
(4.41)
Thus, we can also express the von Mises stress in terms of the stress components in the
xy coordinate system. For plane stress conditions, we have
σe =
(σ x + σ y )2 − 3(σ x σ y − τ 2xy )
(4.42)
119
Two-Dimensional Elasticity
4.4.7.2 Averaged Stresses
Stresses are usually averaged at nodes in FEA software packages to provide more accurate
stress values. This option should be turned off at nodes between two materials or other
geometry discontinuity locations where stress discontinuity does exist.
EXAMPLE 4.2
A square plate with a hole at the center is under a tension load p in x direction as shown
in the figure.
y
p
A
x
B
The dimension of the plate is 10 in. × 10 in., thickness is 0.1 in., and radius of the hole is
1 in. Assume E = 10 × 106 psi, ν = 0.3, and p = 100 psi. Find the maximum stress in the plate.
FE Analysis
This is a plane stress case. From the knowledge of stress concentrations, we should
expect the maximum stresses occur at points A and B on the edge of the hole. Value of
this stress should be around 3p (=300 psi), which is the exact solution for an infinitely
large plate with a circular hole.
We use the ANSYS to do the modeling (meshing) and analysis, using quadratic triangular (T6), linear quadrilateral (Q4), and quadratic quadrilateral (Q8) elements. The
FEM results by using the three different elements are compared and their accuracies and
efficiencies are discussed. One mesh plot and one stress contour plot are shown below.
An FE mesh (T6, 1518 elements)
–8.006
64.925
137.856
210.787
283.719
28.46
101.391
174.322
320.184
247.253
FE stress plot and deformed shape
(T6, 1518 elements)
120
Finite Element Modeling and Simulation with ANSYS Workbench
TABLE 4.1
FEA Stress Results
Elem. Type
Q4
No. of Elem.
506
Total DOFs
Max. σ (psi)
1102
312.42
Q4
3352
7014
322.64
Q4
31,349
64,106
322.38
...
...
T6
T6
...
...
1518
6254
320.18
2562
10,494
321.23
T6
24,516
100,702
322.24
...
...
...
...
Q8
501
3188
320.58
Q8
2167
13,376
321.70
Q8
14,333
88,636
322.24
The stress calculations with several meshes are listed in Table 4.1, along with the
­number of elements and DOFs used.
The converged results are obtained with all three types of elements with the differences in the maximum stress values less than 0.05%. However, Q8 and T6 elements are
more efficient and converge much faster than the Q4 elements which are a linear representation and cannot model curved boundaries accurately. If the required accuracy
is set at 1%, then the mesh with 501 Q8 elements should be sufficient. Note also that
we need to check the deformed shape of the plate for each model to make sure the BCs
are applied correctly. Fewer elements should be enough to achieve the same accuracy
with a better or “smarter” mesh (mapped mesh). We will redo this example in the next
chapter employing the symmetry features of the problem.
4.4.8 General Comments on the 2-D Elements
a.
b.
c.
Know the attributes of each type of elements:
T3 and Q4: Linear displacement, constant strain and stress
T6 and Q8: Quadratic displacement, linear strain and stress
Choose the right type of elements for a given problem:
When in doubt, use higher order elements (T6 or Q8) or a finer mesh.
Avoid elements with large aspect ratios and corner angles (Figure 4.17):
Aspect ratio = Lmax/Lmin where Lmax and Lmin are the largest and smallest characteristic lengths of an element, respectively.
d. Make sure the elements are connected properly:
Do not leave unintended gaps or free elements in FE models (Figure 4.18).
121
Two-Dimensional Elasticity
Elements with distorted shapes
Elements with normal shapes
FIGURE 4.17
Elements with distorted (irregular) and normal (regular) shapes.
A
C
B
D
Improper connections (gaps along AB and CD)
FIGURE 4.18
Unintended gaps in the FE mesh.
4.5 Case Study with ANSYS Workbench
Problem Description: A combination wrench is a convenient tool that is used to apply
torque to loosen or tighten a fastener. The wrench shown below is made of stainless steel
and has a thickness of 3 mm. Determine the maximum deformation and the distribution
of von Mises stresses under the given distributed load and boundary conditions.
p
12
12
Material: Stainless steel
E = 193 GPa
ν = 0.27
10
2.5
7
12.5
10
40
Boundary conditions:
Hexagon on the right end is
fixed on all sides.
Pressure p = 2 MPa
(All units are in millimeters.)
Solution
To solve the problem with ANSYS Workbench, we employ the following steps:
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Wrench.wbpj.”
122
Finite Element Modeling and Simulation with ANSYS Workbench
Step 2: Create a Static Structural (ANSYS) Analysis System
Drag the Static Structural (ANSYS) icon from the Analysis Systems Toolbox window
and drop it inside the highlighted green rectangle in the Project Schematic window to create a standalone static structural analysis system.
Step 3: Add a New Material
Double-click (or right-click and choose Edit) on the Engineering Data cell in the
above Project Schematic to edit or add a material. In the following Engineering
Data interface which replaces the Project Schematic, click the empty box highlighted below and type “Stainless Steel” as name for the new material.
Type ‘stainless steel’
in the empty box.
Select “Stainless Steel” from the Outline window, and double-click Isotropic
Elasticity under Linear Elastic in the leftmost Toolbox window. Enter “193E9”
for Young’s Modulus and “0.27” for Poisson’s Ratio in the bottom center
Two-Dimensional Elasticity
Properties window. Click the Return to Project button to go back to the Project
Schematic.
Step 4: Launch the DesignModeler Program
Ensure Surface Bodies is checked in the Properties of Schematic A3: Geometry window (select Properties from the View drop-down menu to enable display of this
window). Select 2D for Analysis Type in this Properties window. Double-click
the Geometry cell to launch DesignModeler, and select “Millimeter” in the Units
pop-up window.
123
124
Finite Element Modeling and Simulation with ANSYS Workbench
Step 5: Create Surface Sketch
Click on the Sketching tab. Select the Draw toolbox and then Rectangle. Draw a
rectangle in the XY-plane in the Graphics window by first clicking on the Y-axis
and then dragging anywhere to the right. This makes the left side of the rectangle coincide with the Y-axis. Select the Dimensions toolbox to specify one
horizontal (H1) and two vertical dimensions (V2 and V3) as shown below.
Select the Draw toolbox and then Circle. Draw a circle by first clicking on the X-axis
for the center of the circle and then at a corner of the rectangle for a point on
the circle, as shown below. The letter C indicates the center has a coincident
relation with X-axis, and the letter P indicates the cursor is on the corner point.
Select Dimensions and specify a radius of 10 mm for the circle.
Two-Dimensional Elasticity
Draw another circle on the left by first clicking on the X-axis and then at the
upper-left corner of the rectangle. Make sure letters C and P show up as you
draw. Specify a radius dimension of 12.5 mm.
Select the Draw toolbox and then Polygon. Draw a six-sided polygon (set n = 6)
in the Graphics window by first clicking at the center of the right circle. Then,
drag the cursor anywhere inside the circle and click when there is a letter H
showing above the top edge of the polygon. The letter H indicates the edge is
a horizontal line. Select the Dimensions toolbox and specify the polygon side
length as 7 mm.
125
126
Finite Element Modeling and Simulation with ANSYS Workbench
Select the Modify toolbox and then Trim. Click on unwanted line segments and
arcs (a total of four lines and four arcs) in the Graphics window to trim them
away. The trimmed sketch looks like the following.
Select the Draw toolbox and then Rectangle. Draw a rectangle on the left as shown
below. Select the Dimensions toolbox and specify one horizontal dimension
(H7) and two vertical dimensions (V8 and V9) as given below.
Two-Dimensional Elasticity
Select the Modify toolbox and then Trim. Trim away unwanted line segments
and arcs (a total of two lines and two arcs) by clicking on them. The trimmed
sketch looks like the following.
Select the Modify toolbox and then Fillet. Set the radius of fillet as 2.5 mm. Click on
two intersecting edges to define a fillet. Create two fillets as shown below at
the corners of the rectangle.
127
128
Finite Element Modeling and Simulation with ANSYS Workbench
Step 6: Create Surface Body
Switch to the Modeling tab and select Surfaces from Sketches from the drop-down
menu of Concept.
Select Sketch1 from the Tree Outline and click Apply on the Base Objects selection in
the Details of SurfaceSK1. Then click Generate.
Two-Dimensional Elasticity
A Surface Body is now created from the surface sketch. Set the Thickness to 3 mm in
the Details of Surface Body. Exit the DesignModeler.
Step 7: Launch the Static Structural Program
Double-click on the Model cell to launch the Static Structural program. Click on
Geometry in the Outline. In the Details of “Geometry,” the following options are
available for 2D Behavior: plane stress, axisymmetric, plane strain, generalized
plane strain (assuming a finite length in the z direction, as opposed to the infinite value assumed for the standard plane stain option), and by body (allowing you to set 2D behavior options for individual bodies that appear under
Geometry in the outline). Choose Plan Stress as the desired 2D Behavior.
129
130
Finite Element Modeling and Simulation with ANSYS Workbench
In the Details of “Surface Body,” click to the right of the Material Assignment field
and select Stainless Steel from the drop-down context menu.
Step 8: Generate Mesh
Right click on Mesh in the Project Outline tree. Select Insert and then Method from
the context menu. Click on the surface body in the Graphics window, and apply
it to the Geometry selection in the Details of “Automatic Method.”
Two-Dimensional Elasticity
Right click on Mesh in the Project Outline. Select Insert and then Sizing from the
context menu.
In the Details of “Body Sizing,” enter “‘1.5e-3 m” for the Element Size. Click on the
surface body in the Graphics window and apply it to the Geometry selection. In
the Outline of Project, right-click on the Mesh and select Generate Mesh.
131
132
Finite Element Modeling and Simulation with ANSYS Workbench
The following mesh plot will show up.
Step 9: Apply Boundary Conditions
Right-click on Static Structural (A5) and select Insert and then Fixed Support from
the context menu.
133
Two-Dimensional Elasticity
Ctrl-Click all six sides of the polygon as highlighted in the Graphics window. You
must turn on the Edge Selection filter to select a line. Click Apply on the Geometry
selection field in the Details of “Fixed Support.”
Click to turn on the edge
selection filter.
Step 10: Apply Loads
In the Outline, right-click on Static Structural (A5). Choose Insert and then Pressure.
134
Finite Element Modeling and Simulation with ANSYS Workbench
In the Details of “Pressure”, click on the arc as shown below and apply it to the
Geometry selection. The Edge Selection filter must be turned on to allow this
selection. Choose Vector for the Define By field, and enter “2e6” for Magnitude.
Click on the vertical line as shown below and click Apply to confirm on the
Direction selection. Make sure the vector is pointing downward. If not, click the
left or right arrow at the bottom left corner to toggle direction.
Click on the arc to
apply pressure.
Click on this vertical
line to use as a
direction reference.
Click on the left or right arrow to
toggle the vector direction.
Step 11: Retrieve Solution
Insert a Total Deformation item by right-clicking on Solution (A6) in the Outline tree.
Two-Dimensional Elasticity
Insert an Equivalent Stress item by right-clicking on Solution (A6) in the Outline
tree.
Right-click on Solution (A6) in the Outline tree and select Solve. The program will
start to solve the model. After completion, click Total Deformation in the Outline
to review the total deformation results. As shown below, the maximum deformation occurs at the two end tips on the left side of the wrench with a magnitude of 0.138 mm under the assigned pressure load.
135
136
Finite Element Modeling and Simulation with ANSYS Workbench
Select Equivalent Stress in the Outline to review the von Mises stress distribution.
The following figure shows that the maximum von Mises stress is located
at the two entrant corners on the right side of the wrench with a magnitude
of 165.68 MPa, which is below 207 MPa, the yield strength of stainless steel.
Decrease the element size in the Details of “Body Sizing” and solve again to see
if there is any increase in the solution accuracy.
Modeling tips: ANSYS Workbench can work with native geometric models created in
other CAD systems such as SolidWorks, CATIA, or Pro/E. It also allows geometry to be
imported in neutral file formats such as IGES, Parasolid, and SAT.
To import a part created in SolidWorks, for example, you can right-click on the Geometry
cell. Then choose Import Geometry and browse to select the SolidWorks part file.
Next, double-click the Geometry cell to launch DesignModeler. Click Generate in DesignModeler
to load the geometry. Then, choose Surfaces From Faces from the Concept menu.
Browse for the file
you wish to import.
137
Two-Dimensional Elasticity
Click on the front wrench face and apply it to the Faces selection in the Details of
SurfFromFace1. Click Generate to complete the surface creation.
Click on the
front face.
In the Tree Outline, a new Wrench item is now added underneath 2 Parts, 2 Bodies. Rightclick on the first Wrench item in the Outline, and choose Suppress Body in the context menu.
This will hide the solid wrench and display only the new surface model. Click on the new
Wrench item, and enter a “3 mm” for the Thickness in the Details of Surface Body and exit
DesignModeler. Double-click on the Model cell in the Project Schematic to launch the Static.
Structural program. A concept wrench surface is now ready for downstream analysis.
138
Finite Element Modeling and Simulation with ANSYS Workbench
4.6 Summary
The 2-D elasticity equations are reviewed in this chapter and 2-D elements for analyzing
plane stress and plane strain problems are discussed. FE formulations for 2-D stress analysis are introduced. It is emphasized that linear triangular (T3) and linear quadrilateral (Q4)
elements are good for deformation analysis and not accurate for stress analysis. For stress
analysis, quadratic triangular (T6) and quadratic quadrilateral (Q8) elements are recommended. Bad-shaped elements with large aspect ratios and large or small angles should
be avoided in an FE mesh. A wrench model is built and the stress is analyzed in ANSYS
Workbench to show how to conduct a 2-D FEA.
4.7 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1.
2.
3.
4.
Develop the FE formulations for 2-D stress analysis problems.
Create a mesh for a 2-D stress analysis problem and be able to judge the mesh quality.
Apply the load and constraints to the FE model correctly.
Conduct 2-D stress analysis of an FE model using ANSYS Workbench.
PROBLEMS
4.1 List the boundary conditions in Example 4.1.
139
Two-Dimensional Elasticity
4.2 The plate shown below is constrained at the left end and loaded with a linearly
varying pressure load at the right end. Constants E, ν, and thickness t are given.
y
po
h
x
h
L
Suppose we have found the displacement field as follows:
p0
xy ,
Eh
p
v = − 0 ( x 2 + νy 2 )
2Eh
u=
Find:
a. strains in the plate;
b. stresses in the plate;
c. check if or not the equilibrium equations are satisfied by the stresses;
d. check if or not the boundary conditions are satisfied by the solution.
Optional:
Assume E = 10 × 106 psi, ν = 0.3, po = 100 psi, L = 12 in., h = 4 in., and thickness
t = 0.1 in. Use ANSYS Workbench to check your results.
4.3 Derive the shape functions in Equation 4.17 for T3 elements and prove
(Equation 4.18).
4.4 From Equation 4.27, prove det J = x13 y 23 − x23 y13 = 2 A and discuss why “badshaped” elements can cause numerical errors in the FEM.
4.5 The torque arm shown below is a 5 mm thick automotive component made of
structural steel with a Young’s modulus of 200 GPa and a Poisson’s ratio of 0.3.
Using ANSYS Workbench, determine the deformation and von Mises stress distributions under the given load and boundary conditions.
Fy = 400 N
36 cm
4 cm
1.5 cm
4 cm
6 cm
3 cm
Boundary condition:
The left bolt hole is fixed.
2 cm
4 cm
4 cm
Fx = 3000 N
140
Finite Element Modeling and Simulation with ANSYS Workbench
4.6 A bicycle wrench shown below is made of stainless steel with a Young’s modulus
of 193 GPa and a Poisson’s ratio of 0.27. The wrench is 2 mm thick. Using ANSYS
Workbench, determine the location and magnitude of the maximum deformation and the maximum von Mises stress under the given load and boundary
conditions.
F = 50 N
R = 12 mm
+
6 mm
+
25 mm
R = 12 mm
24 mm
+
20 mm
The side lengths of the hexagons
from left to right are 9, 7,
and 5 mm, respectively.
Boundary and load condition:
The smallest hexagon is fixed on
all sides.
A 50 N force is applied on the
top edge of the wrench.
2 mm
4.7 Consider a straight and long hexagonal pipe under internal pressure as shown
below. The pipe is made of stainless steel with a Young’s modulus of 193 GPa
and a Poisson’s ratio of 0.27. Using ANSYS Workbench, determine the maximum
in-plane deformation and the maximum von Mises stress under the given load
and boundary conditions.
30 mm
The radius of the internal
hole is 20 mm.
The internal pressure:
p = 20 MPa.
p
Boundary condition:
The hexagonal pipe is
fixed on the bottom.
4.8 Consider a straight and deep tunnel under external pressure. The tunnel is
made of concrete with a Young’s modulus of 29 GPa and a Poisson’s ratio of 0.15.
Using ANSYS Workbench, determine the maximum in-plane deformation and
the maximum in-plane von Mises stress under the given load and boundary
conditions.
p
p
p
0.8 m
2.4 m
The external pressure:
p = 1MPa.
1.2 m
0.4 m
1.6 m
Boundary condition:
0.4 m
The tunnel is fixed on
the bottom.
141
Two-Dimensional Elasticity
4.9 Using ANSYS Workbench, design a steel shelf bracket. Some dimensions of the
bracket are fixed as shown in the figure. The three holes have the same diameter, which can be changed in increment of 0.2 cm in the design. The goal is
using as less material as possible for the bracket, while supporting the given
distributed load p.
a. For steel, use E = 200 GPa, ν = 0.3, and yield stress σY = 250 MPa.
b. Use a factor of safety = 2.5 for the design.
c. Report the configuration and dimensions of the bracket of your final design.
p = 250 N/cm2
H = 12 cm
+
2 cm
2 cm
2 cm
+
2 cm
Fillet radius: 5 cm
Bracket thickness:
t = 0.2 cm
L = 20 cm
4.10 Similar to the previous problem, design a steel bracket. Some dimensions of the
bracket are fixed as shown in the figure, while others can be changed. The shape
and topology of the bracket can also be changed. The goal of this design is using
least material for the bracket, while supporting the given loads.
a. For steel, use E = 200 GPa, ν = 0.32, and yield stress σY = 250 MPa.
b. Use a safety factor of 3 for the design.
c. Report the configuration and dimensions of the bracket of your final design.
Bolted to a frame
600 N
60 mm
60 mm
Radius of all three holes: R = 10 mm;
thickness: 3 mm.
600 N
200 mm
5
Modeling and Solution Techniques
5.1 Introduction
In this chapter, we discuss several techniques in the modeling and solution options of
using the FEM. Applying these techniques can greatly improve the efficiencies and accuracies of the FEA.
5.2 Symmetry
Symmetry feature of a structure is the first thing one should look into and explore in
the FE modeling and analysis. The model size can be cut almost in half and the solution efficiencies can be improved by several times. A structure possesses symmetry if its
components are arranged in a periodic or reflective manner. Types of symmetries are
(Figure 5.1):
• Reflective (mirror, bilateral) symmetry
• Axisymmetry
• Rotational (cyclic) symmetry
• Translational symmetry
• Others (or combinations of the above)
In the FEM, symmetry properties can be applied to
• Reduce the size of the problems and thus save CPU time, disk space, postprocessing efforts, and so on
• Simplify the modeling task
• Check the FEM results (make sure the results are symmetrical if the geometry and
loading of the structure are symmetrical)
Symmetry properties of a structure should be fully exploited and retained in the FE
model to ensure the efficiency and quality of FE solutions.
143
144
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 5.1
Some examples of symmetry: (a) reflective symmetry (From http://www.thinkingfountain.org/s/­symmetry/
butterflypattern.gif); (b) axisymmetry; (c) rotational symmetry (From http://csdt.rpi.edu/na/pnwb/­
symmetry2a.html); and (d) translational symmetry (From http://library.thinkquest.org/16661/background/
symmetry.1.html).
5.2.1 An Example
For the problem of a plate with a center hole as discussed in Example 4.2 of the previous
chapter, we redo the FE mesh using the symmetry features of the plate. To do this, we first
model just one-quarter of the plate using mapped mesh, and then reflect the model (with
the mesh) twice to obtain the model and mesh for the entire plate, as shown in Figure 5.2.
Only 896 Q8 elements are used in this symmetrical model and the results are comparable
to those in Chapter 4 using more elements with the free mesh. The quarter model can also
be applied in the analysis, if the boundary conditions are also symmetrical about the xz
and yz planes.
In vibration or buckling analysis, however, the symmetry concept should not be used in
the FEA solutions (it is still applicable in the modeling stage), since symmetric structures
often have antisymmetric vibration or buckling modes.
Z
Z
–7.34
29.07
65.48
101.89
138.3
211.12
283.941
174.71
247.531
320.351
FIGURE 5.2
Results using symmetry features for Example 4.2 (mesh and stress contour plots).
Modeling and Solution Techniques
145
5.3 Substructures (Superelements)
Another very useful technique for analyzing very large FEA models of mechanical systems is to apply the concept of substructures or superelements (SEs). Substructuring is a
process of analyzing a large structure as a collection of (natural) components. The FEA
models for these components are called substructures or superelements (SEs). The physical
meaning of a substructure is simply a finite element model of a portion of the structure.
Mathematically, it presents a boundary matrix which is condensed by eliminating the interior points and keeping only the exterior or boundary points of the portion of the structure.
In other words, instead of solving the FEA system of equations once, one can use partitions
of the matrix so that larger models can be solved on relatively smaller computers. More
details of the theory and implementations of the substructures or SEs can be found in the
documentation of the FEA software packages.
Figure 5.3 shows an FEA model of a truck used to conduct the full vehicle static or
dynamic analysis. The entire model can have several millions of DOFs that can be beyond
the capabilities of some computers. Using the substructuring technique, one can build the
FEA model for each subsystem first (such as the cab, chassis, steering system, suspension
system, payload, and so on) and then condense the FEA equations to smaller ones relating
only to DOFs on the interfaces between the subsystems and residing on a residual structure (e.g., the chassis). The condensed system is much smaller than the original system and
can be solved readily.
The advantages of using the substructuring technique are
• Good for large problems (which will otherwise exceed your computer capabilities)
• Less CPU time per run once the SEs have been processed (i.e., matrices have been
condensed and saved)
• Components may be modeled by different groups
• Partial redesign requires only partial reanalysis (reduced cost)
• Efficient for problems with local nonlinearities (such as confined plastic deformations) which can be placed in one SE (residual structure)
• Exact for static deformation and stress analysis
FIGURE 5.3
An FEA model of a truck analyzed using substructures.
146
Finite Element Modeling and Simulation with ANSYS Workbench
The disadvantages of using the substructuring technique are
• Increased overhead for file management
• Increased initial time for setting up the system
• Matrix condensations for dynamic problems introduce new approximations
5.4 Equation Solving
There are two types of solvers used in the FEA for solving the linear systems of algebraic
equations, mainly, the direct methods and iterative methods.
5.4.1 Direct Methods (Gauss Elimination)
• Solution time proportional to NB2 (with N being the dimension of the matrix and
B the bandwidth of the FEA systems)
• Suitable for small to medium problems (with DOFs in the 100,000 range), or slender structures (small bandwidth)
• Easy to handle multiple load cases
5.4.2 Iterative Methods
• Solution time is unknown beforehand
• Reduced storage requirement
• Suitable for large problems, or bulky structures (large bandwidth, converge
faster)
• Need to solve the system again for different load cases
5.4.3 An Example: Gauss Elimination
Solve the following given system of equations:
8
−2
0
−2
4
−3
0 x1 2
−3 x2 = −1 or Ax = b
3 x3 3
(5.1)
Forward elimination:
Form
(1) 8
(2) −2
(3) 0
−2
4
−3
0| 2
|
−3 −1
|
3 | 3
(5.2)
147
Modeling and Solution Techniques
(1) + 4x(2) ⇒ (2):
(2) +
(1) 8
(2) 0
(3) 0
−2
14
−3
0 | 2
|
−12 −2
|
3 | 3
(5.3)
(1) 8
(2) 0
(3) 0
−2
14
0
0 | 2
|
−12 −2
|
2 | 12
(5.4)
14
(3) ⇒ (3)
3
Back substitutions (to obtain the solution):
x3 = 12/2 = 6
1.5
x2 = (−2 + 12x3 )/14 = 5 or x = 5
6
(5.5)
x1 = (2 + 2x2 )/8 = 1.5
5.4.4 An Example: Iterative Method
The Gauss–Seidel method (as an example):
Ax = b (A is symmetric)
(5.6)
or
N
∑a x
ij
j
= bi , i = 1, 2,..., N
j =1
Start with an estimate x(0) of the solution vector and then iterate using the following:
xi( k + 1) =
1
bi −
aii
i −1
∑
or i = 1, 2,..., N
aij x j( k ) , fo
j =i +1
N
aij x j( k + 1) −
j =1
∑
(5.7)
In vector form,
x( k + 1) = A D −1[b − A L x( k + 1) − A LT x( k ) ]
where
A D = ⟨aii⟩ is the diagonal matrix of A,
A L is the lower triangular matrix of A,
(5.8)
148
Finite Element Modeling and Simulation with ANSYS Workbench
such that
A = AD + A L + A LT
(5.9)
Iterations continue until solution x converges, that is
x( k + 1) − x( k )
x( k )
≤ ε
(5.10)
where ε is the tolerance for convergence control.
Iterative solvers with moderate selections of the tolerance are usually much faster than
direct solvers in solving large-scale models. However, for ill-conditioned systems, direct
solvers should be applied to ensure the accuracy of the solutions.
5.5 Nature of Finite Element Solutions
FEA model is a mathematical model of the real structure, based on many approximations.
An observation is made on the FEA models and solutions:
• Real structure—infinite number of nodes (physical points or particles), thus infinite number of DOFs
• FEA model—finite number of nodes, thus finite number of DOFs
In particular, one can argue that the displacement field is controlled (or constrained) by
the values at a limited number of nodes (Figure 5.4).
Therefore, we have the so-called stiffening effect:
• FEA model is stiffer than the real structure.
• In general, displacement results are smaller in magnitudes than the exact values.
Hence, the FEM solution of displacement is a lower bound of the exact solution.
4
Recall that on an element: u = Σ Nα uα
α=1
FIGURE 5.4
Elements in an FEA model.
149
Modeling and Solution Techniques
Δ (Displacement)
Exact solution
FEA solutions
No. of DOFs
FIGURE 5.5
Convergence of FEM solutions with exact solution.
That is, FEA displacement solutions approach the exact solution from below (Figure 5.5),
which can be used to monitor the FEA solutions. However, this is true only for the displacement-based FEA (which is the FEA formulation discussed in this book).
5.6 Convergence of FEA Solutions
As the mesh in an FEA model is “refined” (with smaller and smaller elements) repeatedly,
the FEA solution will converge to the exact solution of the mathematical model of the problem (the model based on beam, plane stress/strain, plate, shell, or 3-D elasticity theories or
assumptions). Several types of refinements have been devised in the FEA, which include:
h-refinement: Reduce the size of the element (“h” refers to the typical size of the
elements).
p-refinement: Increase the order of the polynomials on an element (linear to quadratic, and so on; “p” refers to the highest order in a polynomial).
r-refinement: Rearrange the nodes in the mesh.
hp-refinement: Combination of the h- and p-refinements (to achieve better results).
With any of the above type of refinements, the FEA solutions will converge to the analytical solutions of the mathematical models. Some FEA software can automate the process
of refinements in the FEA solutions to achieve the so-called adaptive solutions.
5.7 Adaptivity (h-, p-, and hp-Methods)
Adaptive FEA represents the future of the FEA applications. With proper error control,
automatic refinements of an FEA mesh can be generated by the program until the converged FEA solutions are obtained. With the adaptive FEA capability, users’ interactions
150
Finite Element Modeling and Simulation with ANSYS Workbench
are reduced, in the sense that a user only needs to provide a good initial mesh for the
model (even this step can be done by the software automatically).
Error estimates are crucial in the adaptive FEA. Interested readers can refer to Reference
[5] for more details. In the following, we introduce one type of the error estimates.
We first define two stress fields:
σ—element by element stress field (discontinuous across elements)
σ*—averaged or smoothed stress field (continuous across elements)
Then, the error stress field can be defined as
σE = σ − σ*
(5.11)
Compute strain energies,
M
U=
∑U ,
i
Ui =
i =1
∑U ,
*
i
U *i =
i =1
∑U
Ei
1
*T
∫ 2σ
−1
σ dV
(5.12)
E −1 σ * dV
(5.13)
Vi
M
UE =
T
Vi
M
U* =
1
∫ 2σ E
, UE i =
i =1
1
∫ 2σ E
T
E
−1
σ E dV
(5.14)
Vi
where M is the total number of elements and Vi is the volume of the element i.
One error indicator—the relative energy error is defined as
UE
η=
U + UE
1/2
. (0 ≤ η ≤ 1)
(5.15)
The indicator η is computed after each FEA solution. Refinement of the FEA model continues until, say
η ≤ 0.05
When this condition is satisfied, we conclude that the converged FE solution is obtained.
Some examples of using different error estimates in the FEA solutions can be found in
Reference [5].
5.8 Case Study with ANSYS Workbench
Problem Description: Garden fountains are popular amenities that are often found at
theme parks and hotels. As a fountain structure is usually an axisymmetric geometry
151
Modeling and Solution Techniques
with axisymmetric loads and support, only a 2-D model, sliced through the 3-D geometry, is needed to correctly predict the deformation of or stress in the structure. The figure
below gives the cross section of an axisymmetric model of a two-tier garden fountain
made of concrete. Determine the maximum deformation and von Mises stress under the
given hydrostatic pressure. Use adaptive meshing to improve solution convergence.
0.06
Material: Concrete
0.12
E = 29 GPa
ν = 0.15
0.66
0.6
Boundary conditions:
0.5
0.06
0.1
0.12
0.8
1
0.2
Bottom edge: fixed.
Left edge: axis of symmetry.
Hydro pressure on dotted edges.
(All units are in meter)
Solution
To solve the problem with ANSYS® Workbench, we employ the following steps:
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Fountain.wbpj”
Step 2: Create a Static Structural (ANSYS) Analysis System
Drag the Static Structural (ANSYS) icon from the Analysis Systems Toolbox window
and drop it inside the highlighted green rectangle in the Project Schematic window to create a standalone static structural analysis system.
152
Finite Element Modeling and Simulation with ANSYS Workbench
Step 3: Add a New Material
Double-click on the Engineering Data cell to add a new material. In the following
Engineering Data interface which replaces the Project Schematic, type “Concrete”
as the name for the new material, and double-click Isotropic Elasticity under
Linear Elastic in the leftmost Toolbox window. Enter “29E9” for Young’s Modulus
and “0.15” for Poisson’s Ratio in the Properties window. Click the Return to
Project button to go back to Project Schematic.
Step 4: Launch the DesignModeler Program
Ensure Surface Bodies is checked in the Properties of Schematic A3: Geometry window (select Properties from the View drop-down menu to enable display of this
window). Select 2D for Analysis Type in this Properties window. Double-click
the Geometry cell to launch DesignModeler, and select “Meter” in the Units popup window.
Modeling and Solution Techniques
Step 5: Create Surface Sketch
Click on the Sketching tab. Select the Draw toolbox and then Line. Draw a closedloop line profile as dimensioned below. Make sure a horizontal constraint (H)
and a vertical constraint (V) appear when drawing a horizontal and a vertical
line, respectively. Use the Fillet tool in the Draw toolbox to create line fillets
with a radius of 0.1 m as shown below.
Step 6: Create Surface Body
Switch to the Modeling tab and choose Surfaces from Sketches from the Concept
menu.
Select Sketch1 from the Tree Outline shown below, and apply it to the Base Objects
selection in the Details of SurfaceSK1. Then click Generate.
A Surface Body will be created from the surface sketch. Exit the DesignModeler.
153
154
Finite Element Modeling and Simulation with ANSYS Workbench
Step 7: Launch the Static Structural Program
Double-click on the Model cell to launch the Static Structural program. Click on
Geometry in the Outline. In the Details of “Geometry,” choose axisymmetric for
2D Behavior.
In the Details of “Surface Body,” click to the right of the Material Assignment field
and select Concrete from the drop-down context menu.
Step 8: Generate Mesh
Right click on Mesh in the Outline. Select Insert and then Sizing from the context
menu. In the Details of “Body Sizing,” enter “0.05 m” for Element Size. Click on
the surface body in the Graphics window and apply it to the Geometry selection.
Right click on Mesh. Select Insert and then Method. In the Details of “Automatic
Method,” click on the surface body, and apply it to the Geometry. Select Triangles
for Method. This will make use of triangular elements for the mesh generation.
Modeling and Solution Techniques
In the Details of “Mesh,” choose Dropped for the Element Midside Nodes field. This
will specify the use of linear elements in the mesh. Note that linear triangular
elements are employed here to show the convergence of linear FEA approximate solutions; they are in general not recommended for stress analysis. Rightclick on Mesh and select Generate Mesh.
Step 9: Apply Boundary Conditions
Right-click on Static Structural (A5). Choose Insert and then Fixed Support from the
context menu. Apply this support to the horizontal line at the bottom.
Right-click on Static Structural (A5). Choose Insert and then Frictionless Support
from the context menu. Apply this support to the leftmost vertical line (center
line of the fountain). The frictionless support prevents the line from moving
or deforming in the normal direction, and thus is equivalent to a symmetry
condition.
Step 10: Apply Loads
In the Project Outline, right-click on Static Structural (A5). Choose Insert and then
Hydrostatic Pressure. The hydrostatic load simulates pressure due to fluid
weight.
In the Details of “Hydrostatic Pressure,” ctrl-click the horizontal line and the adjacent line fillet shown below, and apply the two edges to the Geometry selection. Enter 1000 kg/m3 for Fluid Density. Change the Define By selection to
Components, and enter 9.8 m/s2 for the Y Component of Hydrostatic Acceleration.
Enter 0.68 m for the X Coordinate and 0.76 m for the Y Coordinate for the Free
Surface Location. The location corresponds to the upper endpoint of the line
fillet.
155
156
Finite Element Modeling and Simulation with ANSYS Workbench
Ctrl-click the line and the adjacent line
fillet for hydro-pressure load.
Insert another Hydrostatic Pressure load, and apply the pressure to the line and its
two adjacent line fillets as shown below.
Ctrl-click the line the adjustment two line
fillets for hydro-pressure load.
Modeling and Solution Techniques
Step 11: Retrieve Solution
Insert a Total Deformation item by right-clicking on Solution (A6) in the Outline tree.
Right-click on the Total Deformation in the Outline tree, and choose Insert then
Convergence.
In the Details of “Convergence,” enter 1% for the Allowable Change field.
In the Details of “Solution (A6),” set Max Refinement Loops as 10, and Refinement
Depth as 1. The refinement depth controls the aggressiveness of the mesh
refinement; it has a range from 0 to 3 with a larger number indicating more
aggressive refinement.
Insert an Equivalent Stress item by right-clicking on Solution (A6) in the Outline
tree.
157
158
Finite Element Modeling and Simulation with ANSYS Workbench
Right-click on Solution (A6) in the Outline tree and choose Solve. The program
will start to iterate the solution until the difference between two consecutive
iterations is less than 1% or the maximum number of mesh refinement loops
reaches 10. After completion, click on Convergence in the Outline to review the
convergence curve. The resulting maximum deformations at different mesh
iterations are also recorded in the table below the curve.
Click on Total Deformation in the Outline to review the converged deformation
results.
Modeling and Solution Techniques
159
Click on Equivalent Stress in the Outline to review the stress results.
Modeling tips: A model may be subjected to body forces such as gravitational or radial
centrifugal/inertia forces, in addition to the hydrostatic pressure load. To consider such
forces, the density of the structure’s material needs to be given as an input, and the forces
are typically calculated as follows:
f r =ρ r ω 2 ,
Equivalent radial centrifugal/inertial force
f z = −ρ g ,
Graavitational force
where ρ is the mass density and g the gravitational acceleration (=9.8 m/s2).
Take the following steps to add body forces to the fountain model. First, double-click on
Density in Physical Properties Toolbox. Enter 2.38e3 for Density in the Properties of Outline Row
3: Concrete. Click on the Return to Project button.
160
Finite Element Modeling and Simulation with ANSYS Workbench
Next, refresh the Model cell of Project Schematic after the above change is made on
the Engineering Data cell. Double-click on the Model cell to launch the Static Structural
Program.
In the Project Outline shown below, right-click on Static Structural (A5). Choose Insert and
then Standard Earth Gravity.
Right-click on Static Structural (A5) and choose Insert then Rotational Velocity.
Modeling and Solution Techniques
161
In the Details of “Rotational Velocity,” change Define By to Components. Enter 5 rad/s for Y
Component.
Right-click on Solution (A6) in the Outline tree, and select Solve to update the model results.
The new deformation and stress results are shown below. Both the maximum deformation
and the maximum von Mises stress values are shown to be slightly increased, as compared
to the results considering only the hydrostatic pressure load.
5.9 Summary
In this chapter, we discussed a few modeling techniques and options related to FEA solutions. For static analysis of symmetrical structures, the symmetry features should be
explored in both the modeling (meshing) stage and the solution stage (if the BCs are symmetrical as well). Substructuring or using SEs is a useful technique for solving large-scale
162
Finite Element Modeling and Simulation with ANSYS Workbench
problems with constrained computing resources. Convergence of the FEA solutions is the
important goal in FEA and should be monitored by using the error estimates and employing the adaptive FEA capabilities in the software. The showcase using ANSYS Workbench is
introduced to show how to apply the various techniques introduced in this chapter.
5.10 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1. Apply symmetry in meshing and analysis for a symmetrical structure.
2. Apply axisymmetric model and elements when the structure and loads are
axisymmetric.
3. Select proper equation solvers (direct or iterative) for the FE model at hand.
4. Understand the sources of errors in FEA and the error indicators.
5. Know how to refine a mesh using h-refinement or p-refinement.
6. Apply the various techniques in the FEA of a model using ANSYS Workbench.
PROBLEMS
5.1 Consider the two plate structures shown below. The structures are symmetrically loaded on the x–y plane. Use symmetry to construct appropriate finite
element models for the structures, indicating load and support conditions.
q
q
F
F
5.2 For a tapered bar shown below, study the deformation and stresses in the bar
with a 2-D model using axisymmetric elements and a 1-D model using bar
elements in ANSYS Workbench. Assume R1 = 1 m, R 2 = 0.5 m, L = 5 m, force
F = 3000 N, Young’s modulus E = 200 GPa, and Poisson’s ratio ν = 0.3. The bar is
fixed at the left end. Compare the results from the 2-D and 1-D models.
F
2R1
R2
L
163
Modeling and Solution Techniques
5.3 An open cylinder shown below has an inner radius a = 1 m, outer radius b = 1.1 m,
length L = 6 m, and is applied with a pressure load p = 10 GPa on the inner surface. The outer surface of the cylinder is prevented from radial deformation, and
the cylinder is free to deform along the length direction. Compute the stresses
in the cylinder using a 2-D axisymmetric model in ANSYS Workbench. Assume
Young’s modulus E = 200 GPa and Poisson’s ratio ν = 0.3.
a
b
5.4
L
The rotating part sketched below is made of steel with Young’s modulus
E = 200 GPa, Poisson’s ratio ν = 0.3, and mass density ρ = 7850 kg/m3. Assume
that the part is rotating at a speed of 100 RPM about the z axis. Ignore the gravitational force. Compute the stresses in the part using an axisymmetric model in
ANSYS Workbench.
z
ω
50 mm
30 mm
12 mm
5 mm 8 mm
30 mm
16 mm
5.5 Determine the deformation and stress distributions for a solid cylinder under
the influence of the gravity force and radial centrifugal force using ANSYS
Workbench. The cylinder is made of aluminum with Young’s modulus E = 73 GPa,
Poisson’s ratio ν = 0.3, and mass density ρ = 2800 kg/m3. Assume that the
part is rotating at a speed of 50 rad/s about its central axis. Radius = 10 mm,
height = 20 mm, angular velocity around cylinder central axis = 50 rad/s.
ω
10 mm
20 mm
164
Finite Element Modeling and Simulation with ANSYS Workbench
5.6 Suppose that we need to find out the in-plane effective modulus of a composite
reinforced with long fibers aligned in the z-direction and distributed uniformly.
A 2-D elasticity model shown below can be used for this study with ANSYS
Workbench. The effective modulus can be estimated by using the formula
Eeff = σ x ( ave )/ε x ( ave ), where the averaged stress and strain are evaluated along the
vertical edge on the right side of the model. Assume for the matrix E = 10 GPa,
ν = 0.35, and for the fibers E = 100 GPa, ν = 0.3. The unit cell has a dimension of
1 × 1 μm2 and the radius of the fibers is 0.2 μm.
Start with 1 × 1 cell, 2 × 2 cells, 3 × 3 cells, … and keep increasing the number
of the cells as you can. Report the value of the effective Young’s modulus of the
composite in the x direction. Employ symmetry features of the model in generating the meshes for your analysis.
y
p
x
One unit cell
5.7 Suppose that a “meshed panel” will be used in a design in order to reduce the
weight. For this purpose, we need to find out the in-plane effective modulus of
this panel in the x- or y-direction. A sample piece of the panel similar to the one
shown below can be used for this study. Employ symmetry and study the effects
of the numbers of cells used in the model on the effective modulus of the panel
in the x-direction.
y
a
b
c
p
One cell
x
Assume the panel is made of aluminum with E = 70 GPa, ν = 0.35, a = 10 mm,
b = 5 mm, c = 1.5 mm, and thickness t = 1 mm.
6
Plate and Shell Analyses
6.1 Introduction
Many structure members can be categorized as plates and shells, which are extensions of
the 1-D straight beams and curved beams to 2-D cases, respectively. Examples of structures that can be modeled as plates are shear walls, floor panels, and shelves, while those
that can be modeled as shells include shell structures in nature (such as sea shells and egg
shells), various containers, pipes, tanks, roofs of buildings (such as the superdome), and
bodies of cars, boats, and aircrafts. Figure 6.1 shows an airplane (Boeing 787) that is constructed mainly using plate and shell structure members.
The advantages of using plate and shell structures are their light weight, superior loadcarrying capabilities, and sometimes, simply their artistic appeals.
6.2 Review of Plate Theory
Plates are flat surfaces applied with lateral loading, with bending behaviors dominating
the structural response. Shells are structures which span over curved surfaces; they carry
both membrane and bending forces under lateral loading. The thickness t of a plate or a
shell is much smaller than the other dimensions of the structure. Theories related to plates
and shells are briefly reviewed below.
6.2.1 Force and Stress Relations in Plates
Consider an infinitesimally small element in a plate under lateral loading. The internal
shear forces and bending moments acting on the element, and the induced stresses are
shown below in Figure 6.2 and Figure 6.3, respectively.
The induced stresses are related to the internal moments and forces in the following
manner:
Bending moments (per unit length):
Mx =
∫
t /2
− t /2
σ x z dz , ( N ⋅ m/m)
(6.1)
165
166
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 6.1
An airplane made of numerous plate and shell structures.
z
∆y
y
∆x
My
q(x, y)
Mxy
Qy
t
Mx
x
Qx
Mxy
Mid-surface
FIGURE 6.2
Forces and moments acting on an infinitesimally small element in a plate.
z
τyz
σy
τxy
x
τxz
σx
τxy
FIGURE 6.3
Stresses acting on the infinitesimally small element in the plate.
y
167
Plate and Shell Analyses
My =
∫
t /2
∫
t /2
− t /2
σ y z dz , ( N ⋅ m/m)
(6.2)
τ xy z dz , ( N ⋅ m/m)
(6.3)
Twisting moment (per unit length):
Mxy =
− t /2
Shear Forces (per unit length):
Qx =
Qy =
∫
t /2
∫
t /2
τ xz dz , ( N/m)
(6.4)
τ yz dz , ( N/m)
(6.5)
6 My
6 Mx
, (σ y )max = ± 2
2
t
t
(6.6)
− t /2
− t /2
Maximum bending stresses:
(σ x )max = ±
Similar to the beam model, there is no bending stress at the mid-surface and the maximum/minimum stresses are always at z = ±t/2.
6.2.2 Thin Plate Theory (Kirchhoff Plate Theory)
The thin plate theory is based on assumptions that a straight line normal to the midsurface remains straight and normal to the deflected mid-surface after loading (Figure
6.4); that is
γxz = γyz = 0
(Negligible transverse shear deformations)
z
∂w
∂x
w
x
FIGURE 6.4
Deflection and rotation after loading of a plate according to Kirchhoff plate theory.
168
Finite Element Modeling and Simulation with ANSYS Workbench
The displacement, strains, and stresses can be written in terms of the main variable—
deflection w = w(x, y), as follows:
Displacement:
w = w( x , y ), (deflection)
∂w
∂x
∂w
v = −z
∂y
u = −z
(6.7)
Strains:
∂2 w
∂x2
∂2 w
εy = − z 2
∂y
εx = − z
γ xy = − 2z
(6.8)
∂2 w
∂x∂ y
Note that there is no stretch of the mid-surface caused by the deflection of the plate.
Stresses (plane stress state):
σx
1
E
ν
σy =
2
τ 1 − ν 0
xy
0
εx
0
εy
(1 − ν)/2 γ xy
ν
1
0
(6.9)
or,
σx
1
E
σ
=
−
z
ν
y
1 − ν2
τ
0
xy
ν
1
0
∂ 2w
2
∂x
0 2
∂ w
0 2
∂y
(1 − ν) 2
∂w
∂x∂y
(6.10)
The following equation governs the equilibrium of the plate in the z-direction:
D∇ 4 w = q( x , y )
where
∂4
∂4
∂4
∇4 ≡ 4 + 2 2 2 +
∂x ∂ y
∂ y 4
∂x
(6.11)
169
Plate and Shell Analyses
D=
Et 3
(the bending rigidity of the plate)
12(1 − ν2 )
(6.12)
q = lateral distributed load (force per unit area).
Note that Equation 6.11 is of analogous form with the following 1-D equation for straight
beam:
EI
d4w
= q( x)
dx 4
(6.13)
The fourth-order partial differential equation, given in Equation 6.11 and in terms of
the deflection w(x,y), needs to be solved under certain given boundary conditions. Typical
boundary conditions for plate bending include:
∂w
=0
∂n
Clamped: w = 0,
(6.14)
Simply supported: w = 0, Mn = 0
(6.15)
Free: Qn = 0, Mn = 0
(6.16)
where n is the normal direction of the boundary (Figure 6.5). Note that the given values in
the boundary conditions shown above can be nonzero values as well.
6.2.2.1 Example: A Thin Plate
A square plate (Figure 6.6) with four edges clamped or hinged, and under a uniform load
q or a concentrated force P at the center C.
For this simple geometry, Equation 6.11 with the boundary condition of Equation 6.14 or
6.15 can be solved analytically. The maximum deflections are given in Table 6.1 for the four
different cases. These values can be used to verify the FEA solutions.
s
n
Boundary
FIGURE 6.5
The boundary of a plate.
170
Finite Element Modeling and Simulation with ANSYS Workbench
z
Given: E, t, and ν = 0.3
C
x
y
L
L
FIGURE 6.6
A square plate.
TABLE 6.1
Deflection at the Center (with D = Et3/(12(1 − v 2)))
Clamped
Under uniform load q
Under concentrated force P
Simply Supported
0.00126 qL4/D
0.00560 PL2/D
0.00406 qL4/D
0.0116 PL2/D
6.2.3 Thick Plate Theory (Mindlin Plate Theory)
If the thickness t of a plate is not small, for example, when t/L ≥ 1/10 (L = a characteristic
dimension of the plate main surface), then the thick plate theory by Mindlin should be
applied. The theory accounts for the angle changes within a cross section, that is
γ xz ≠ 0,
γ yz ≠ 0
(transverse shear deformations)
This means that a line which is normal to the mid-surface before the deformation will
not be so after the deformation (Figure 6.7).
The new independent variables θx and θy are rotation angles of a line, which is normal to
the mid-surface before the deformation, about x- and y-axis, respectively.
The following new relations hold:
u = z θ y , v = − zθ x
z
(6.17)
θy ≠ –
∂w
∂x
w
∂w
∂x
FIGURE 6.7
Displacement and rotation based on the Mindlin thick plate theory.
x
171
Plate and Shell Analyses
and
εx = z
γ xz
∂θ y
,
∂x
εy = − z
∂w
=
+ θy ,
∂x
γ yz
∂θ x
,
∂y
∂θ y ∂θ x
γ xy = z
−
,
∂x
∂y
∂w
=
− θx
∂y
(6.18)
Note that if we impose the conditions (or assumptions) that
γ xz =
∂w
∂w
+ θ y = 0, γ yz =
− θx = 0
∂x
∂y
(6.19)
then we can recover the relations applied in the thin plate theory.
The governing equations and boundary conditions can be established for thick plates
based on the above assumptions, with the three main variables involved being w(x, y),
θx(x, y), and θy(x, y).
6.2.4 Shell Theory
Unlike the plate models, where only bending forces exist, there are two types of forces in
shells, that is, membrane forces (in plane forces) and bending forces (out of plane forces)
(Figures 6.8 and 6.9).
6.2.4.1 Shell Example: A Cylindrical Container
Similar to the plate theories, there are two types of theories for modeling shells, namely
thin shell theory and thick shell theory. Shell theories are the most complicated ones to
formulate and analyze in mechanics. Many of the contributions were made by Russian
scientists in the 1940s and 1950s, due to the need to develop new aircrafts and other lightweight structures. Interested readers can refer to Reference [6] for in-depth studies on this
subject. These theoretical works have laid the foundations for the development of various
finite elements for analyzing shell structures.
FIGURE 6.8
Forces and moments in a shell structure member.
172
Finite Element Modeling and Simulation with ANSYS Workbench
p
p
Internal forces:
p
p
Membrane stresses
dominate
FIGURE 6.9
Internal forces acting in a cylinder under internal pressure load.
6.3 Modeling of Plates and Shells
Plates or shells can be modeled as flat or curved surfaces in space, with the thickness t
assigned as a parameter (Figure 6.10). Discretization of the surfaces will involve the use
of plate or shell elements, with the quality of the surface mesh improving with decreasing
element size.
Figure 6.11 gives an example of a stamping part analyzed using shell elements. The
bracket has a uniform thickness and is fixed at the four bolt hole positions. A load is
applied through a pin passing through the two holes in the lower part of the bracket. Note
that one layer of elements on the edge of each hole has been masked in the stress contour
plot (Figure 6.11b), due to inaccurate stress results near the constraint locations. To reduce
the true stress levels in the bracket, the thickness can be changed, the shape of the bracket
can be modified, and the model is remeshed and reanalyzed, all of which are very easy to
carry out with the shell elements.
In many cases, however, the plate and shell models may not be adequate for analyzing a
structure member, even if it is considered thin. For example, the structure component has
a nonuniform thickness (turbine blades, vessels with stiffeners, thin layered structures,
etc.), see Figure 6.12, or has a crack for which detailed stress analysis is needed. In such
t
FIGURE 6.10
A shell structure member and its mathematical representation.
Plate and Shell Analyses
173
FIGURE 6.11
Stress analysis of a bracket using shell elements: (a) the FEA model; (b) stress contour plot.
FIGURE 6.12
Cases in which shell elements are not adequate: (a) casting parts; (b) parts with nonuniform thickness. 3-D solid
elements should be applied in such cases.
cases, one should turn to 3-D elasticity theory and apply solid elements which will be discussed in the next chapter.
6.4 Formulation of the Plate and Shell Elements
Several types of finite elements suitable for the analyses of plates and shells are introduced
below.
6.4.1 Kirchhoff Plate Elements
The following four-node quadrilateral element has only bending capabilities. It has three
DOFs at each node, that is, w , ∂ w/∂y, and ∂ w/∂y (Figure 6.13).
174
Finite Element Modeling and Simulation with ANSYS Workbench
z
y
4
Mid-surface
3
x
2
1
w1 ,
∂w
∂w
,
∂x 1 ∂y
t
w2 ,
1
∂w
∂w
,
∂x 2 ∂y
2
FIGURE 6.13
A four-node quadrilateral element with 3 DOFs at each node.
On each element, the deflection w(x, y) is represented by
4
w( x , y ) =
∑ N w + N
i
i
xi
i =1
∂w
∂w
+ N yi
∂x i
∂ y i
(6.20)
where Ni, Nxi, and Nyi are shape functions. This is an incompatible element [7]. The stiffness
matrix is still of the form
k=
∫ B EB dV
T
(6.21)
V
where B is the strain–displacement matrix and E Young’s modulus (stress–strain)
matrix.
6.4.2 Mindlin Plate Elements
The following two quadrilateral elements are Mindlin types with only bending capabilities (Figure 6.14). There are three DOFs at each node, that is, w, θx, and θy.
On each element, the displacement and rotations are represented by
n
w( x , y ) =
∑N w
i
i
i =1
n
θx (x, y) =
∑N θ
i
xi
i =1
(6.22)
n
θ y (x, y) =
∑N θ
i
yi
i =1
For these elements, there are three independent fields within each element. Deflection
w(x, y) is linear for Q4, and quadratic for Q8.
175
Plate and Shell Analyses
z
z
y
4
2
t
3
8
x
1
7
4
3
y
1
6
t
x
2
5
FIGURE 6.14
Four- and eight-node quadrilateral plate elements.
6.4.3 Discrete Kirchhoff Elements
This is a triangular element with only bending capabilities. First, start with a six-node
triangular element (Figure 6.15). There are 5 DOFs at each corner node (w , ∂w/∂ x , ∂ w/ ∂y ,
θ x , θ y ) 2 DOFs (θ x , θ y ) at each mid-node, and a total of 21 DOFs for the six-node element.
Then, impose conditions γxz = γyz = 0, and so on, at selected nodes to reduce the DOFs
(using relations in Equation 6.18), to obtain the following discrete Kirchhoff triangular
(DKT) element (Figure 6.16).
For the three-node DKT element shown above, there are 3 DOFs at each node, that is,
w, θ x ( = ∂ w/∂ x ), and θ y ( = ∂ w/∂ y ) , and a total of 9 DOFs for the element. Note that w(x, y)
is incompatible for DKT elements [7]; however, its convergence is faster (w is cubic along
each edge) and it is efficient.
6.4.4 Flat Shell Elements
A flat shell element can be developed by superimposing a plane stress element to a plate
element (Figure 6.17).
y
z
3
4
6
2
1
5
x
FIGURE 6.15
A six-node triangular element with 5 DOFs at each corner node and 2 DOFs at each mid-node.
z
1
y
3
2
FIGURE 6.16
Discrete Kirchhoff triangular element with 3 DOFs at each node.
x
176
Finite Element Modeling and Simulation with ANSYS Workbench
+
Plane stress element
(membrane)
Plate element
(bending)
Flat shell element
FIGURE 6.17
Combination of plane stress and plate bending elements yields a flat shell element.
w
v
θx
u
θy
FIGURE 6.18
Q4 or Q8 shell elements.
This is analogous to the combination of a bar element and a simple beam element to
yield a general beam element for modeling curved beams. A flat shell element, with the
DOFs labeled at a typical node i, is shown in Figure 6.18.
6.4.5 Curved Shell Elements
Curved shell elements are based on the various shell theories. They are the most general
shell elements (flat shell and plate elements are subsets). An eight-node curved shell element is illustrated in Figure 6.19, with the DOFs labeled at a typical node i. Formulations
of the shell elements are relatively complicated. They are not discussed here and detailed
derivations are available in References [7–9].
θz
i
w
v
i
θy
FIGURE 6.19
An eight-node curved shell element and the DOFs at a typical node i.
u
θx
177
Plate and Shell Analyses
6.5 Case Studies with ANSYS Workbench
Problem Description: Vases are decorative pieces that can be of any artistic shapes. The
figure below gives the dimensions of a flower vase made of glass. Assume that the vase
has a uniform thickness of 4 mm. The water level reaches 100 mm below the opening of
the vase. Determine the maximum deformation and von Mises stress in the vase under
the hydrostatic pressure.
80 mm
E
280 mm
D
C
B
A
Material: Glass
E = 70 GPa
ν = 0.17
Boundary conditions:
Bottom surface: fixed.
Coordinates of construction points:
A: (50, 0, and 0 mm)
B: (90, 40, and 0 mm)
C: (60, 120, and 0 mm)
D: (40, 180, and 0 mm)
E: (80, 280, and 0 mm)
50 mm
Solution
To solve the problem with ANSYS® Workbench, we employ the following steps:
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Vase.wbpj.”
Step 2: Create a Static Structural (ANSYS) Analysis System
Drag the Static Structural (ANSYS) icon from the Analysis Systems Toolbox window
and drop it inside the highlighted green rectangle in the Project Schematic window to create a standalone static structural analysis system.
178
Finite Element Modeling and Simulation with ANSYS Workbench
Step 3: Add a New Material
Double-click on the Engineering Data cell to add a new material. In the following
Engineering Data interface which replaces the Project Schematic, type “Glass” as
the name for the new material, and double-click Isotropic Elasticity under Linear
Elastic in the leftmost Toolbox window. Enter “70E9” for Young’s Modulus and
“0.17” for Poisson’s Ratio in the Properties window. Click the Return to Project
button to go back to Project Schematic.
Step 4: Launch the DesignModeler Program
Ensure Surface Bodies is checked in the Properties of Schematic A3: Geometry window (select Properties from the View drop-down menu to enable display of this
window). Choose 3D as the Analysis Type in this Properties window. Doubleclick the Geometry cell to launch DesignModeler, and select “Millimeter” in the
Units pop-up window.
179
Plate and Shell Analyses
Step 5: Create a Profile Sketch
Click on the Sketching tab. Select the Draw toolbox and then Construction Point.
Draw five construction points A, B, C, D, and E as shown below. Draw a horizontal line from the origin to point A. Next, draw a spline passing through
points A, B, C, D, and E; right-click at the last construction point and choose
Open End from the context menu to finish the spline creation.
E
D
C
B
A
Step 6: Create a Surface Body
Switch to the Modeling tab and click on the Revolve feature.
The default Base Object is set as Sketch1 in the Details of Revolve1. Click on the y-axis
in the Graphics window and apply it to the Axis selection. Click Generate.
Click on the
y-axis and
apply it to
the axis
selection of
the revolve
feature.
180
Finite Element Modeling and Simulation with ANSYS Workbench
A Surface Body will be created from revolving the profile sketch. Exit the
DesignModeler program.
Step 7: Launch the Static Structural Program
Double-click on the Model cell to launch the Static Structural program. Click on
Surface Body under Geometry in the Outline tree. In the Details of “Surface Body,”
set Thickness as 4e-3 m. Click to the right of the Material Assignment field and
select Glass from the drop-down menu.
Step 8: Generate Mesh
Right click on Mesh in the Project Outline. Select Insert and then Sizing from the
context menu. In the Details of “Face Sizing,” enter “2e-3 m” for the Element Size.
Click on the side wall and the bottom surface of the vase in the Graphics window and apply the two faces to the Geometry selection.
Plate and Shell Analyses
Right-click on Mesh and select Generate Mesh.
Step 9: Apply Boundary Conditions
Right-click on Static Structural (A5). Choose Insert and then Fixed Support from the
context menu. Apply this support to the bottom surface of the vase.
Step 10: Apply Loads
Right-click on Static Structural (A5). Choose Insert and then Hydrostatic Pressure.
In the Details of “Hydrostatic Pressure,” Ctrl-click the side wall and the bottom surface of the vase, and apply the two faces to the Geometry selection.
181
182
Finite Element Modeling and Simulation with ANSYS Workbench
Change the Shell Face to Bottom, because the hydrostatic pressure is applied
to the inside of the shell surface. Enter 1000 kg/m3 for Fluid Density. Change
the Define By selection to Components, and enter 9.8 m/s2 for the Y Component
of Hydrostatic Acceleration. Enter 180e-3 m in the field of Y Coordinate of the Free
Surface Location.
Step 11: Retrieve Solution
Insert a Total Deformation item by right-clicking on Solution (A6) in the Outline tree.
Insert an Equivalent Stress item by right-clicking on Solution (A6) in the Outline
tree.
Right-click on Solution (A6) in the Outline tree and select Solve.
Click on Total Deformation in the Outline to review the deformation results.
Plate and Shell Analyses
183
Click on Equivalent Stress in the Outline to review the stress results.
Modeling tips: Surface bodies can be created from planar 2D sketches, or by revolving,
extruding, or sweeping lines or curves. In many cases, they can also be created from solid
models, by using the Mid-Surface tool provided in Workbench.
For example, to create surface bodies from the solid model of a display shelf given below,
select Mid-Surface from the drop-down menu of Tools.
184
Finite Element Modeling and Simulation with ANSYS Workbench
In the “Details of MidSurf1,” change the Selection Method to Automatic. The shelf panel has
a thickness of 4 mm; set the Minimum Threshhold as 3 mm, and the Maximum Threshhold
as 5 mm. In the field of Find Face Pairs Now?, choose Yes from the drop-down menu. Click
Generate.
Eight surface bodies will be generated based on automatic face pair detection, as shown
in the following figure. Exit DesignModeler, and the new model is now ready for subsequent simulation with plate or shell elements.
185
Plate and Shell Analyses
6.6 Summary
In this chapter, the main aspects of the plate and shell theories and the plate and shell elements used for analyzing plate and shell structures are discussed. Plates and shells can
be regarded as the extensions of the beam elements from 1-D line elements to 2-D surface
elements. Plates are usually applied in modeling flat thin structure members, while the
shells are applied in modeling curved thin structure members. In applying the plate and
shell elements, one should keep in mind the assumptions used in the development of these
types of elements. In cases where these assumptions are no longer valid, one should turn
to general 3-D theories and solid elements. A showcase using shell elements in modeling a
vase is introduced using ANSYS Workbench.
6.7 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1. Understand the assumptions used in the plate and shell theories.
2. Understand the behaviors of the plate and shell elements.
3. Know when not to use plate and shell elements in stress analysis (e.g., when plate
or shell structures have nonuniform thickness or small features).
4. Create FEA models using plate and shell elements for deformation and stress analysis using ANSYS Workbench.
PROBLEMS
6.1 The square plate shown below has the following dimensions and material constants: L = 1 m, t = 0.02 m, E = 30 × 106 N/m2, and ν = 0.3. Use plate elements to
determine the maximum displacement and von Mises stress in the slab in the
following cases: (a) The four edges are clamped and the slab is under a uniform
load q = 1 N/m2. (b) The four edges are clamped and the slab is under a concentrated force P = 1 N at the center C. (c) The four edges are hinged and the slab is
under a uniform load q = 1 N/m2. (d) The four edges are hinged and the slab is
under a concentrated force P = 1 N at the center C. Verify the FEA solutions with
the analytical solutions given in the Table 6.1.
z
C
x
L
y
L
186
Finite Element Modeling and Simulation with ANSYS Workbench
6.2 The roof structure shown below is loaded by its own weight with q = 90 lbf/
in2. The dimensions and material constants are: R = 25 in., L = 50 in., t = 0.25 in.,
E = 432 × 106 psi, and ν = 0.0. The two straight edges are free, while the two
curved edges have a “diaphragm” support (meaning that x and y DOFs are constrained, but z (along the length axis) and all rotational DOFs are unconstrained).
Use shell elements to find the maximum displacement and von Mises stress in
the structure. Verify your results (note that the value of the analytical solution
for the displacement at the mid-point A of the straight edge is 0.3024 in.).
L
q
A
R
80°
6.3 The pinched cylinder shown below is loaded by a force F = 1 N. The dimensions
and material constants are: R = 300 mm, L = 600 mm, t = 3 mm, E = 3 × 106 N/mm2,
and ν = 0.3. Circular ends have diaphragm support (meaning that y and z DOFs
are constrained, but x (along the length axis) and all rotational DOFs are unconstrained). Use shell elements to find the maximum displacement and von Mises
stress in the structure. Verify your results (note that the value of the analytical
solution for the displacement at the mid-point A is 0.1825 × 10−4 mm).
L/2
L/2
F
A
R
F
6.4 The pinched hemisphere shown below has radial loads F = 2 N at the equator. The
dimensions and material constants are: R = 10 m, t = 0.04 m, E = 68.25 × 106 N/m2,
and ν = 0.3. The hemisphere has a free edge, and is restrained only against rigid
body motion. Use shell elements to find the maximum displacement and von
Mises stress in the structure. Verify your results (note that the value of the analytical solution for the displacement at the point A is 0.0924 m).
F
R
F
A
F
F
187
Plate and Shell Analyses
6.5 The twisted strip shown below has a load F = 10−6 N. The dimensions and
material constants are: b = 1.1 m, L = 12 m, t = 0.0032 m, E = 29 × 106 N/m2, and
ν = 0.22. The strip has a 90° twist over length L and is cantilevered. Use shell elements to find the maximum displacement and von Mises stress in the structure.
F
b
A
L
6.6 The twisted strip shown below has a load F = 10−6 N. The dimensions and
material constants are: b = 1.1 m, L = 12 m, t = 0.0032 m, E = 29 × 106 N/m2,
and ν = 0.22. The strip has a 90° twist over length L and is cantilevered. Use
shell elements to find the maximum displacement and von Mises stress in the
structure.
b
A
L
F
6.7 Consider a glass cup placed on a table, as shown in the figure. Using shell elements, find the maximum displacement and von Mises stress in the cup when
the cup is applied with a pressure load of 10 N/mm2 on the inner wall. Assume
that the cup has a uniform thickness of 4 mm, E = 70 GPa, and ν = 0.17.
150 mm
40 mm
30 mm
188
Finite Element Modeling and Simulation with ANSYS Workbench
6.8 Consider a glass fish tank placed on a table, as shown in the figure. Using shell
elements, find the maximum displacement and von Mises stress in the tank
when the tank is full of water. Assume that the tank has a uniform thickness of
4 mm, E = 70 GPa, and ν = 0.17.
60 mm
100 mm
60 mm
6.9 A fuel tank, with a total length = 5 m, diameter = 1 m, and thickness = 0.01 m, is
shown below. Assume Young’s modulus E = 200 GPa and Poisson’s ratio ν = 0.3.
Using shell elements, find the deformation and stresses when the tank is applied
with an internal pressure p = 1 MPa and placed on the ground.
6.10 A bookshelf shown below is placed on the ground. Assume Young’s modulus
E = 13.1 GPa, Poisson’s ratio ν = 0.29, and the shelf thickness t = 6 mm. Using
shell elements, find the deformation and stresses when the bookshelf is applied
with a downward uniform pressure p = 500 N/m2 on the entire surface of the
bookshelf middle layer.
20 cm
20 cm
30 cm
50 cm
30 cm
30 cm
7
Three-Dimensional Elasticity
7.1 Introduction
Engineering designs involve 3-D structures that cannot be adequately represented using
1-D or 2-D models. Solid elements based on 3-D elasticity [10,11] are the most general elements for stress analysis when the simplified bar, beam, plane stress/strain, plate/shell
elements are no longer valid or accurate. In general, 3-D structural analysis is one of the
most important and powerful ways of providing insight into the behavior of an engineering design. In this chapter, we will review the elasticity equations for 3-D and then
discuss a few types of finite elements commonly used for 3-D stress analysis. Several different types of supports, loads, and contact constraints will be introduced for 3-D structural modeling, followed by a case study on predicting the deformation and stresses in
an assembly structure using ANSYS® Workbench.
7.2 Review of Theory of Elasticity
The state of stress at a point in a 3-D elastic body is shown in Figure 7.1.
In vector form, the six independent stress components determining the state of stress
can be written as
σx
σ
y
σ z
σ = {σ} =
τ xy
τ yz
τ zx
, or [σ ij ]
(7.1)
189
190
Finite Element Modeling and Simulation with ANSYS Workbench
y
x
z
y, v
σy
τyx
τyz
τxy
τzy
σx
τzx
σz
τxz
x, u
z, w
FIGURE 7.1
State of stress at a point in a 3-D elastic body.
Similarly, the six independent strain components in 3-D can be expressed as
εε
ε
ε = {ε } =
γ
γ
γ
x
y
z
xy
yz
zx
,
(7.2)
or [ε ij ]
7.2.1 Stress–Strain Relation
The stress–strain relation in 3-D is given by
1 − v
v
σx
σ
v
y
σ z
E
0
=
τ
v
v
(
1
+
)(
1
−
2
)
xy
τ yz
0
τ
zx
0
v
v
v
0
0
0
1 − 2v
2
0
0
0
1−v
v
1−v
0
0
0
0
0
1 − 2v
2
0
0
0
0
0
εx
εy
0 ε z
γ xy
0 γ yz
γ
1 − 2v zx
2
0
0
0
(7.3)
191
Three-Dimensional Elasticity
Or in a matrix form:
σ = Eε
7.2.2 Displacement
The displacement field can be described as
u( x , y , z) u1
u = v( x , y , z) = u2
w( x , y , z) u
3
(7.4)
7.2.3 Strain–Displacement Relation
Strain field is related to the displacement field as given below
∂u
∂v
∂w
, εy =
, εz =
,
∂x
∂y
∂z
∂v
∂u
∂w
∂v
∂u
∂w
=
+
, γ yz =
+
, γ xz =
+
∂x
∂y
∂y
∂z
∂z
∂x
εx =
γ xy
(7.5)
These six equations can be written in the following index or tensor form:
ε ij =
(
)
∂u j
1 ∂ui
+
, i , j = 1, 2, 3
2 ∂x j
∂x i
Or simply,
ε ij =
1
(ui , j + u j ,i ) (Tensor notation)
2
7.2.4 Equilibrium Equations
The stresses and body force vector f at each point satisfy the following three equilibrium
equations for elastostatic problems:
∂σ x ∂τ xy ∂τ xz
+ f x = 0,
+
+
∂x
∂y
∂z
∂τ yx ∂σ y ∂τ yz
+
+
+ f y = 0,
∂x
∂y
∂z
∂τ zx ∂ τ zy ∂σ z
+
+
+ fz = 0
∂y
∂z
∂x
(7.6)
192
Finite Element Modeling and Simulation with ANSYS Workbench
p
Γσ
n
Γ (= Γu + Γσ)
Γu
FIGURE 7.2
The boundary of a 3-D elastic domain.
Or in index or tensor notation:
σ ij , j + f i = 0
7.2.5 Boundary Conditions
At each point on the boundary Γ and in each direction, either displacement or traction
(stress on the boundary) should be given, that is
ui = ui ,
ti = ti ,
on Γ u (specified displacement);
on Γ σ (specified traction)
(7.7)
in which the barred quantities denote given values, and the traction (stress on a surface) is
defined by ti = σ ij n j or in a matrix form:
t x σ x
ty = τ xy
t τ
z xz
τ xy
σy
τ yz
τ xz nx
τ yz ny
σ z nz
with n being the normal (Figure 7.2).
7.2.6 Stress Analysis
For 3-D stress analysis, one needs to solve equations in Equations 7.3, 7.5, and 7.6 under
the BCs in Equation 7.7 in order to obtain the stress, strain, and displacement fields (15
equations for 15 unknowns for a 3-D problem). Analytical solutions are often difficult to
find and thus numerical methods such as the FEA are often applied in 3-D stress analysis.
7.3 Modeling of 3-D Elastic Structures
3-D stress analysis using solid elements is one of the most challenging tasks in FEA. In
the following, practical considerations in 3-D FEA modeling such as mesh and BCs are
discussed. The types of contact used in assembly analyses are presented.
Three-Dimensional Elasticity
193
7.3.1 Mesh Discretization
Meshing structures with complicated geometries can be very tedious and time-­
consuming. Great care should be taken to ensure that the FEA mesh is in good quality
(e.g., with no distorted elements). Computing cost is another factor. For structures with
stress concentrations, large FEA models are often needed, which can run hours or days. A
good CAE engineer should be able to decide where to apply a fine mesh and where not to,
in order to strike a balance between the computational cost and accuracy for an FEA task.
Figure 7.3 shows an example of an FEA model using solid elements of a drag link in a
car. Although the structure has a slender shape, it has a bended angle and holes. 3-D solid
elements are needed for the stress analysis in this case. Great care is taken in meshing this
part, where quadratic elements are used for better accuracy in the stress analysis. Buckling
analysis may also be conducted for slender structures when they are under compressions.
More information about buckling analysis using FEA can be found in Chapter 12.
Figure 7.4 shows a 3-D FEA of a gear coupling which is applied to transmit power
through two aligned rotating shafts. Contact stresses and failure modes are to be determined based on detailed 3-D FE models. This analysis requires the use of nonlinear FEA
options, which are readily available now in almost all FEA software packages.
7.3.2 Boundary Conditions: Supports
A number of support types are available for 3-D structural analysis. The following list
includes three common support conditions.
• Fixed support: Prevent the geometry entity from moving or deforming.
• Frictionless support: Prevent the face geometry from moving or deforming in the
normal direction relative to the face.
• Cylindrical support: Prevent the cylindrical face from moving or deforming in any
combination of radial, axial, or tangential directions relative to the cylinder.
FIGURE 7.3
FEA for a drag link: (a) the model; (b) mesh for the right end; (c) stress distribution due to tension loads applied
at the two ends.
194
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 7.4
Analysis of a gear coupling: (a) ring gear; (b) hub gear; (c) high-contact stresses in the gear teeth obtained using
nonlinear FEA.
Note that there are six rigid-body motions for 3-D bodies: three translations and three
rotations. These rigid-body motions (causes of singularity of the system of equations) must
be removed from the FEA model for stress analysis to ensure the accuracy of the analysis.
On the other hand, over constraints can also cause inaccurate or unwarranted results. For
more information on the support conditions, please refer to the Mechanical User Guide of
the ANSYS help documents [12].
7.3.3 Boundary Conditions: Loads
The types of structural loads that can be encountered in 3-D analyses include force, moment,
pressure, bearing load, and so on. Inertia loads such as acceleration, standard earth gravity, or rotational velocity may have nontrivial effect on structures’ stress behaviors as well.
Other loading types such as thermal, electric, or magnetic loads can also be involved, but
are less common. For more information on the structural loads, please see Reference [12].
7.3.4 Assembly Analysis: Contacts
For assembly analysis, contact conditions are needed to describe how different contacting
bodies can move relative to one another.
The following types of contact are available for assembly analysis, as listed in Reference
[12]:
• Bonded: This is the default configuration. Bonded regions can be considered as
glued together, allowing no sliding or separation between the contacting regions.
For many applications, bonded contact is sufficient for stress calculations between
bolted or welded parts in assemblies.
195
Three-Dimensional Elasticity
• No separation: This contact type allows frictionless sliding along the contact faces,
but separation of faces in contact is not allowed.
• Frictionless: This contact model allows free sliding, assuming a zero coefficient of
friction. Gap can form in between regions in contact.
• Rough: This model assumes an infinite friction coefficient between the bodies in
contact. No sliding can occur.
• Frictional: This model allows bodies in contact to slide relative to each other, once
an equivalent shear stress up to a certain magnitude is exceeded.
In most cases, contact regions can be automatically detected and generated in the FEA
program. They can also be manually modified, if needed.
7.4 Formulation of Solid Elements
In this section, we will first summarize the FEA formulation for 3-D elasticity problems,
which are straightforward extensions of the FEA formulations for 1-D bar and 2-D elasticity problems. We will then use an example of linear hexahedral (eight-node brick) element
to examine the element formulation in detail.
7.4.1 General Formulation
As in the FEA formulations for 1-D and 2-D problems, we first interpolate the displacement
fields within a 3-D element using shape functions Ni:
N
u=
∑N u ,
i
i
i =1
N
v=
∑N v ,
(7.8)
i
i
i =1
N
w=
∑N w
i
i
i =1
in which ui, vi, and wi are nodal values of the displacement on the element, and N is the
number of nodes on that element. In matrix form, we have:
uv
w
uv
w
1
1
u
=
v
w
( 3 × 1)
N0
0
1
0
N1
0
0
0
N1
N2
0
0
0
N2
0
0
0
N2
(3 × 3 N )
1
2
2
2
( 3 N × 1)
(7.9)
196
Finite Element Modeling and Simulation with ANSYS Workbench
Or in a matrix form:
u=Nd
Using relations given in Equations 7.5 and 7.8, we can derive the strain vector to obtain:
ε=Bd
in which B is the matrix relating the nodal displacement vector d to the strain vector ε.
Note that the dimensions of the B matrix are 6 × 3N.
Once the B matrix is found, one can apply the following familiar expression to determine the stiffness matrix for the element:
∫
k = BT EB dv
(7.10)
v
The dimensions of the stiffness matrix k are 3N × 3N. A numerical quadrature is often
needed to evaluate the above integration, which can be expensive if the number of nodes
is large, such as for higher-order elements.
7.4.2 Typical Solid Element Types
We can classify the type of elements for 3-D problems as follows (Figure 7.5) according to
their shapes and the orders of the shape functions constructed on the elements:
Tetrahedron:
Linear (four nodes)
Quadratic (10 nodes)
Hexahedron (brick):
Linear (eight nodes)
Quadratic (20 nodes)
Pentahedron:
Linear (six nodes)
FIGURE 7.5
Different types of 3-D solid elements.
Quadratic (15 nodes)
197
Three-Dimensional Elasticity
Whenever possible, one should try to apply higher-order (quadratic) elements, such as
10-node tetrahedron and 20-node brick elements for 3-D stress analysis. Avoid using the
linear, especially the four-node tetrahedron elements in 3-D stress analysis, because they
are inaccurate for such purposes. However, it is fine to use them for deformation analysis
or in vibration analysis (see Chapter 8).
7.4.3 Formulation of a Linear Hexahedral Element Type
Displacement Field in the Element:
8
u=
∑
8
N i ui , v =
i =1
∑
8
N i vi , w =
i =1
∑N w
i
i
(7.11)
i =1
Shape Functions:
1
(1 − ξ)(1 − η)(1 − ζ),
8
1
N 2 (ξ , η, ζ) = (1 + ξ)(1 − η)(1 − ζ),
8
1
N 3 (ξ , η, ζ) = (1 + ξ)(1 + η)(1 − ζ),
8
1
N 8 (ξ , η, ζ) = (1 − ξ)(1 + η)(1 + ζ)
8
N1(ξ , η,ζ) =
(7.12)
Note that we have the following relations for the shape functions:
N i (ξ j , η j , ζ j ) = δ ij , i , j = 1, 2,… , 8.
8
∑ N (ξ, η, ζ) = 1
i
i =1
Coordinate Transformation (Mapping):
8
x=
∑
8
N i xi , y =
i =1
∑
i =1
8
Ni yi , z =
∑N z
i i
(7.13)
i =1
That is, the same shape functions are used for the element geometry as for the displacement field. This kind of element is called an isoparametric element. The transformation
between (ξ, η, ζ ) and (x, y, z) described by Equation 7.13 is called isoparametric mapping
(see Figure 7.6).
Jacobian Matrix:
∂u
∂ξ
∂u
=
∂η
∂u
∂ζ
∂x
∂ξ
∂x
∂η
∂x
∂ζ
∂z ∂u
∂ ξ ∂x
∂z ∂u
∂ η ∂y
∂y
∂z ∂u
∂ζ
∂ζ ∂z
≡ J Jacobian matrix
∂y
∂ξ
∂y
∂η
(7.14)
198
Finite Element Modeling and Simulation with ANSYS Workbench
3
4
1
8
2
7
6
5
x
Mapping (xyz ξηζ)
(–1≤ ξ,η,ζ ≤ 1)
η
(–1, 1, –1) 4
3 (1, 1, –1)
(–1, 1, 1) 8
7 (1, 1, 1)
o
ξ
2 (1, –1, –1)
(–1, –1, –1) 1
(–1, –1, 1) 5
ζ
6 (1, –1, 1)
FIGURE 7.6
Mapping an element to the natural coordinate system.
Inverting this relation, we have:
∂u
∂u
∂ξ
∂x
∂u
∂u
−1 ∂u
=
= J , with
∂ξ
∂y
∂η
∂u
∂u
∂z
∂ζ
8
∂N i
∑ ∂ξ
ui and so on
(7.15)
i =1
and similarly for v and w. These relations lead to the following expression for the strain:
∂u
∂x
∂v
ε x ∂y
ε
y ∂w
ε z ∂z
ε= =
= use ( 6.15) = Bd
∂
v
∂
u
γ xy
+
γ yz ∂x ∂y
∂w ∂v
γ zx
+
∂z
∂y
∂u ∂w
+
∂x
∂z
199
Three-Dimensional Elasticity
where d is the nodal displacement vector, that is:
ε = Bd
(7.16)
Strain energy is evaluated as
U=
1
1
σ T ε dV =
(Eε)T ε dV
2
2
∫
∫
V
V
1 T
ε E ε dV
=
2
∫
V
=
1 T T
d B E B dV d
2
V
∫
(7.17)
That is, the element stiffness matrix is
∫
k = BT EB dV
(7.18)
V
In ξηξ coordinates:
dV = (det J)dξd ηdζ
(7.19)
Therefore,
1 1 1
k=
∫ ∫ ∫ B EB(det J) dξ dη dζ
T
−1 −1 −1
(7.20)
It is easy to verify that the dimensions of this stiffness matrix is 24 × 24.
Stresses:
To compute the stresses within an element, one uses the following relation once the
nodal displacement vector is known for that element:
σ = E ε = EBd
Stresses are evaluated at selected points (Gaussian points or nodes) on each element.
Stress values at the nodes are often discontinuous and less accurate. Averaging of the
stresses from surrounding elements around a node is often employed to smooth the stress
field results.
The von Mises stress for 3-D problems is given by
σ e = σVM =
1
( σ 1 − σ 2 )2 + ( σ 2 − σ 3 ) 2 + ( σ 3 − σ 1 ) 2
2
where σ1, σ2, and σ3 are the three principal stresses.
(7.21)
200
Finite Element Modeling and Simulation with ANSYS Workbench
pA/3
pA/12
p
Surface area = A
Nodal forces for 20-node
brick element
FIGURE 7.7
Equivalent nodal forces on a 20-node brick element for a pressure load p.
7.4.4 Treatment of Distributed Loads
Distributed loads need to be converted into nodal forces using the equivalent energy concept as discussed in earlier chapters. Figure 7.7 shows the result of a pressure load converted into nodal forces for a 20-node hexahedron element. Note the direction of the forces
at the four corner nodes, which is not intuitive at all.
7.5 Case Studies with ANSYS Workbench
Problem Description: A base stand assembly includes a base, a holder, and a pin, as
shown in the following figure. The stand assembly is made of structural steel. Assume a
no-separation condition for all contact regions. Determine the deformation and von Mises
stress distributions of the assembly under the given load and boundary conditions.
Holder
Pin
Material: Structural steel (E= 200 GPa, ν = 0.3)
Boundary conditions:
The bottom faces of the leg base are fixed.
A downward force of 1 kN is applied to the holder’s top face.
Geometry construction:
The bottom of the hub base is 35 mm above the ground level.
Base
The holder is 36 mm tall, 18 mm of which is in contact with
the hub base.
R3.000
R50.000
8.000
R6.000
60.000
R40.000
R10.000
Sketch of the leg base (extrude
5 mm on both sides)
Sketch of the hub base (extrude
35 mm on one side)
Sketch of the pin (extrude
15 mm on both sides)
Three-Dimensional Elasticity
Solution
To solve the problem with ANSYS Workbench, we employ the following steps:
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Assembly.wbpj.”
Step 2: Create a Static Structural (ANSYS) Analysis System
Drag the Static Structural (ANSYS) icon from the Analysis Systems Toolbox window
and drop it inside the highlighted green rectangle in the Project Schematic window to create a standalone static structural analysis system.
Step 3: Launch the DesignModeler Program
Double-click the Geometry cell to launch DesignModeler, and select “Millimeter” in
the Units pop-up window.
Step 4: Create the Base Geometry
Click on the Sketching tab. Select the Draw toolbox and then Line and Arc by Center.
Draw a sketch with two lines and two arcs on the XY Plane, as shown below.
An entity named Sketch1 will be shown underneath XY Plane of the model’s
Tree Outline.
201
202
Finite Element Modeling and Simulation with ANSYS Workbench
Extrude Sketch1 to create a solid body. In Details of Extrude1, change Direction to
Both-Symmetric, and enter 5 mm for FD1, Depth. Click on Generate.
From the drop-down menu of Create, click on Body Operation.
In the Details of BodyOp1, change the Type to Rotate. Click on the solid body in the
Graphics window and apply it to the selection of Bodies. Change Preserve Bodies
Three-Dimensional Elasticity
to Yes. Click on the Y-axis in the Graphics window and apply it to the Axis
Selection. Set the FD9, Angle to 120°. Click on Generate.
From the drop-down menu of Create, click on Body Operation. In the Details of
BodyOp2, change the Type to Rotate. Click on the original solid body in the
Graphics window and apply it to the selection of Bodies. Change Preserve Bodies
to Yes. Click on the Y-axis in the Graphics window and apply it to the Axis
Selection. Set the FD9, Angle to −120°. Click on Generate.
203
204
Finite Element Modeling and Simulation with ANSYS Workbench
Click on New Plane to create the Plane4. In the Details of Plane4, Change the Base
Plane to ZXPlane, and Transform1 (RMB) to Offset Z. Enter 35 mm for FD1,
Value1. Click on Generate.
Click on the new
Plane button.
Click on Look At Face/Plane/Sketch. Draw two concentric circles as shown below,
with a radius given as 6 mm and 10 mm, respectively. Click on Generate to finish
creation of Sketch2.
Click on Look At
Face/Plane/Sketch.
205
Three-Dimensional Elasticity
Extrude Sketch2 along the normal direction for an FD1, Depth of 35 mm, as shown
below. Click on Generate. The new Extrude2 will be combined with the other
three parts into a single solid body.
Step 5: Create the Holder Geometry
Click on New Plane to create the Plane5. In the Details of Plane5, change the Type to
From Face. Click on the annulus face from the Graphics window and apply it to
the selection of Base Face. Click on Generate.
Click on the
New Plane button.
Click on the
annulus face.
206
Finite Element Modeling and Simulation with ANSYS Workbench
Create a New Sketch on Plane5. Draw a circle, the same size as the smaller of the
two concentric circles. Click on Generate. A Sketch3 will be created underneath
Plane5 in the Tree Outline.
Click on the new
sketch button.
Draw a circle here.
Click on Extrude. In the Details of Extrude3, change the Direction to Both- Symmetric,
and set the FD1, Depth as 18 mm. Click on Generate.
Three-Dimensional Elasticity
Create a New Sketch on the YZPlane. Draw a circle with a radius of 3 mm, centered
60 mm to the right of the origin, as shown below. Click on Generate to finish the
creation of Sketch4.
Extrude Sketch4. In the Details of Extrude4, change Operation to Cut Material, and
the Direction to Both- Symmetric. Use the default 30 mm for FD1, Depth. Click on
Generate.
207
208
Finite Element Modeling and Simulation with ANSYS Workbench
Click on Freeze from the drop-down menu of Tools. Freezing a body and then
using a Create Slice can help physically separate merged bodies to create a multipart assembly.
Then click on Slice from the drop-down menu of Create.
209
Three-Dimensional Elasticity
In the Detailed of Slice1, Change the Slice Type to Slice by Surface. Click on the cylindrical surface in the Graphics window and apply it to the selection of Target
Face. Click on Generate.
Click on the surface.
A two-part assembly will be created, as shown below. Right-click on the solid
bodies from the Tree Outline to rename them as Holder and Base, respectively.
210
Finite Element Modeling and Simulation with ANSYS Workbench
Step 6: Create the Pin Geometry
Extrude Sketch4. In the Details of Extrude5, change Operation to Add Material, and
the Direction to Both- Symmetric. Set the FD1, Depth as 15 mm.
Right Click on Solid and rename it as Pin in the Tree Outline. A three-part assembly
is now created.
Step 7: Launch the Static Structural Program
Double-click on the Model cell to launch the Static Structural program. ANSYS
Workbench automatically detects contacts between parts of an assembly, whether
Three-Dimensional Elasticity
it is created in DesignModeler or imported from neutral geometry files. Click on
Contacts under Connections in the Outline tree. In the Details of “Contact Regions,”
change the Type to No Separation for all three contact regions. No separation
contact prevents gap forming between the contact regions; a small amount of
frictionless sliding may occur along the contact faces.
Step 8: Generate Mesh
Right click on Mesh in the Project Outline. Select Insert and then Sizing from the
context menu. In the Details of “Body Sizing,” enter “2.5e-3 m” for the Element
Size. Ctrl-click on the three parts in the Graphics window and apply the bodies
to the Geometry selection.
211
212
Finite Element Modeling and Simulation with ANSYS Workbench
Right-click on Mesh and select Generate Mesh.
Step 9: Apply Boundary Conditions
Right-click on Static Structural (A5). Choose Insert and then Fixed Support from the
context menu. Ctrl-click on the three faces highlighted below and apply them
to the Geometry selection in the Details of “Fixed Support.”
Three-Dimensional Elasticity
Step 10: Apply Loads
Right-click on Static Structural (A5). Choose Insert and then Force. In the Details
of “Force,” apply the circular face shown below to the Geometry selection.
Change the Define By selection to Components, and enter − 1000 N for the Y
Component.
Step 11: Retrieve Solution
Insert a Total Deformation item by right-clicking on Solution (A6) in the Outline tree.
Insert an Equivalent Stress item by right-clicking on Solution (A6) in the Outline.
Right-click on Solution (A6) in the Outline and select Solve.
Click on Total Deformation in the Outline to review the deformation results.
213
214
Finite Element Modeling and Simulation with ANSYS Workbench
Click on Equivalent Stress in the Outline to review the stress results.
To view deformation of the pin, insert a Total Deformation item by right-clicking
on Solution (A6) in the Outline tree. Click on the pin body in the Graphics
window and apply it to the selection of Geometry in the Details of “Total
Deformation 2.”
Three-Dimensional Elasticity
To view stress in the pin, insert an Equivalent Stress item by right-clicking on
Solution (A6) in the Outline tree. Click on the pin body in the Graphics window
and apply it to the selection of Geometry in the Details of “Equivalent Stress 2.”
Repeat the last two steps to add a Total Deformation 3 and an Equivalent Stress 3 to
the Outline. Select the holder body in the Graphics window for the Geometry in
both Details of “Total Deformation 3” and Details of “Equivalent Stress 3.”
Right-click on Solution (A6) in the Outline and select Evaluate All Results. Click on
Total Deformation 2 and Total Deformation 3 in the Outline to review the deformation of the pin and the holder, respectively.
215
216
Finite Element Modeling and Simulation with ANSYS Workbench
Click on Equivalent Stress 2 and Equivalent Stress 3 in the Outline to review the von
Mises stress of the pin and the holder, respectively.
Modeling tips: It is in general a good idea to take advantage of symmetry whenever possible to simplify calculations. The base stand assembly is symmetric about the XY-plane.
To create a half-symmetry model, open “Assembly.wbpj” and double-click on the Geometry
cell in the Project Schematic. In the drop-down menu of Tools of the launched DesignModeler,
select Symmetry, as shown below. An item named Symmetry1 will be added to the Tree
Outline.
Three-Dimensional Elasticity
217
In the Details of Symmetry1, click on XYPlane in the Tree Outline and apply it to the selection of Symmetry Plane1. Click on Generate. A half-model will be created as shown below.
Exit DesignModeler.
Double-click on Model cell in the Project Schematic to launch the Static Structural–
Mechanical program. Click Yes on the popup menu to allow reading the modified upstream
data. Note that a Symmetry Region is now added to the Project Outline as shown below.
218
Finite Element Modeling and Simulation with ANSYS Workbench
Right-click on Mesh in the Outline and select Generate Mesh.
In the half-model, the magnitude of the force applied to a face should be divided by two.
Change the Y Component of the force from −1000N to −500N in the Details of “Force.”
Three-Dimensional Elasticity
219
The frictionless support is equivalent to a symmetry condition. Right-click on Static
Structural (A5). Choose Insert and then Frictionless Support from the context menu. In the
Details of “Frictionless Support,” Ctrl-click the seven faces in the Graphics window as shown
below, and apply them to the Geometry selection field. Note that this step is especially
important for imported geometry with cut faces at the symmetry plane, because the program cannot automatically detect symmetry regions as in the case of geometry created
internally.
Right-click on Static Structural (A5) and select Solve. Click on the deformation and stress
results under Solution (A6) and compare them with results from the full model.
220
Finite Element Modeling and Simulation with ANSYS Workbench
7.6 Summary
In this chapter, we discussed the 3-D solid elements for elasticity problems, that is, general
3-D deformation and stress analyses. Solid elements are the most accurate elements and
should be applied when the bar, beam, plane stress/strain, and plate/shell elements are no
longer valid or accurate. For stress concentration problems, higher-order solid elements,
such as 10-node tetrahedron or 20-node hexahedron (brick) elements, should be employed
in the FEA. For solids having symmetrical features, symmetric FEA models can be more
effective and efficient.
7.7 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1.
2.
3.
4.
Understand the FE formulations for 3-D stress analysis
Know the behaviors of the 3-D (tetrahedron and hexahedron) elements
Create quality mesh over solids for 3-D stress analysis
Perform detailed stress analysis of 3-D structures using ANSYS Workbench
PROBLEMS
7.1
For a tapered bar shown below, study the deformation and stresses in the bar with
a 3-D model using solid elements and a 1-D model using 1-D bar elements. Assume
R1 = 1 m, R 2 = 0.5 m, L = 5 m, force F = 3000 N, Young’s modulus E = 200 GPa, and
Poisson’s ratio ν = 0.3. The bar is fixed at the left end. Compare the FEA results
between the 3-D and 1-D models.
2R1
R2
F
L
7.2
An open cylinder shown below has an inner radius a = 1 m, outer radius b = 1.1 m,
length L = 10 m, and is applied with a pressure load p = 10 GPa on the inner surface. Compute the stresses in the cylinder using shell and solid models. Compare
the FEA results based on these models. Assume Young’s modulus E = 200 GPa and
Poisson’s ratio ν = 0.3.
a
b
L
221
Three-Dimensional Elasticity
7.3
For the rotating part sketched below, assume that it is made of steel with Young’s
modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and mass density ρ = 7850 kg/m3.
Assume that the part is rotating at a speed of 100 RPM about the z axis. Ignore
the gravitational force. Compute the deformation in the part using the FEA with a
full 3-D model. Compare the 3-D results with the axisymmetric modeling results
obtained in Problem 5.4.
z
ω
12 mm
8 mm
30 mm
7.4
50 mm
30 mm
5 mm
16 mm
The following figure shows a simple bracket made of structural steel with Young’s
modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The bracket has a cross-section
profile shown below and an extrusion depth of 12 cm. A through-hole of diameter
1.5 cm is centered on the bracket’s flat ends, one on each side, as shown in the figure. Suppose that the bracket is fixed on the bottom, and lifted by a pressure load
of 1000 N/cm2 from the inside surface of the top flat portion. Compute the deformation and stresses in the bracket using the FEA.
All dimensions are in
centimeters.
10.000
1.500
1.500
R3.000
1.500
8.000
1.500
12.000
8.000
1.500
222
7.5
Finite Element Modeling and Simulation with ANSYS Workbench
The following figure shows a lifting bracket made of structural steel with Young’s
modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The base portion of the bracket
is 20 cm long, 12 cm wide, and 3 cm deep. The fillet radius at the base corners is
1.75 cm. Four through-holes each of radius 1 cm are located symmetrically both
length wise and width wise in the base, with the hole centers being 12 cm apart by
6 cm apart. The upper portion of the base has a profile sketch shown below and is
3 cm thick. Suppose that the bracket is fixed on the bottom, and lifted by a bearing load of 20 kN through the bolt hole located in the upper portion. Compute the
deformation and stresses in the bracket using the FEA.
R3.000
R1.500
All dimensions are in
centimeters.
3.000
3.000
8.000
6.000
6.000
7.6 The following figure shows a bracket made of structural steel with Young’s
modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The rectangular base of the
bracket is 12 cm long, 6 cm wide, and 3 cm deep. The midsection has an overall
extrusion depth of 1.6 cm and a profile sketch shown below. The hollow cylinder
has an inner radius of 1 cm, an outer radius of 2 cm, and a height of 4 cm. The
central axis of the cylinder is 12.5 cm away from the leftmost end of the bracket.
Suppose that the bracket is fixed on the left end, and is applied by a pressure
load of 60 N/cm2 on the inclined face of the midsection. Compute the deformation and stresses in the bracket using the FEA.
All dimensions are in
centimeters.
6.000
8.000
4.000
223
Three-Dimensional Elasticity
7.7
The following figure shows an S-shaped wooden block with Young’s modulus
E = 70 K N/mm2 and Poisson’s ratio ν = 0.3. The block has a uniform thickness of
6 mm and a depth of 8 mm. It is fixed on the bottom and subjected to a surface
pressure of 1 N/mm2 on the top. Compute the deformation and stresses in the
block using the FEA.
All dimensions are in
centimeters.
6.000
6.000
6.000
6.000
6.000
24.000
7.8
The following figure shows an L-shaped bracket made of structural steel with
Young’s modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The 4 cm thick L-shape
with a cross-section profile shown in the figure is rigidly connected to a rectangular base of size 3 cm × 8 cm × 1 cm. Suppose that the bracket is fixed on the right
end at the base. A uniform pressure of 100 N/cm2 is applied on the protruding
surface of the L-shape as shown below. Compute the deformation and stresses in
the bracket using the FEA.
All dimensions are in
centimeters.
1.000
R1.000
3.000
1.000
8.000
224
7.9
Finite Element Modeling and Simulation with ANSYS Workbench
The following two-bar assembly consists of four components made of structural steel with Young’s modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and density
ρ = 7850 kg/m3. As shown in the figure below, the two bars are identical in size.
Each bar has a diameter of 2 cm and a length of 16 cm. The two identical end blocks
each has a thickness of 2 cm and a fillet radius of 1 cm at the round corners. The distance of the gap in between the two blocks is 8 cm. Suppose that the two blocks are
fixed on the bottom. The bars are pulled by a pressure load of 100 N/cm2 in opposite
directions as shown below. Compute the stresses in the assembly using the FEA.
All dimensions are in centimeters.
centimeters
Use bonded contact for all contact regions.
2.000
5.000
R1.000
6.000
10.000
7.10 The following figure shows a half model of a clevis assembly consisting of a yoke,
a pin, and a u-shape. The assembly components are made of structural steel with
Young’s modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The profile sketches of
the yoke and the u-shape are shown in the figure below. Both components have an
extrusion depth of 5 cm. In the half model, the pin has a diameter of 2 cm and a
length of 8 cm, and is centered 2.5 cm away from both the front and the side faces
of the u-shape. Suppose that the yoke is fixed on the left end, and the u-shape is
pulled by a pressure load of 100 N/cm2 on the right end. Compute the stresses in
the assembly using the FEA.
1.500
All dimensions are in centimeters.
Use no separation for all contact regions.
12.000
2.000
R1.500
R1.500
16.000
3.000
1.500
2.000
8.000
2.000
R1.000
2.500
3.000
2.500
8
Structural Vibration and Dynamics
8.1 Introduction
Vibration and dynamics are fundamental subjects in engineering. Serious problems may
arise from vibration when a structure is not carefully designed for its dynamic integrity.
Vibration can cause malfunction or break down of machines that exhibit unbalance or misalignment. It can also lead to massive engineering failures such as the collapse of a bridge.
Simulation plays an important role in our ability to understand a structure’s dynamic
behavior. Through modeling, the dynamic characteristics of a structure can be captured
and improved before being put into actual use.
In this chapter, we first review the basic equations and their solutions for structural
vibration and dynamic analysis. Then, we discuss the FEA formulations for solving vibration and dynamic responses. Guidelines in modeling and solving such problems are provided, along with a case study in ANSYS® Workbench.
A structure vibrates about an equilibrium position when excited by periodic or arbitrary inputs. There are three main types of problems for structural vibration and dynamic
analyses:
• Modal analysis (f(t) = 0)
• Frequency response analysis (f(t) = F sin ωt)
• Transient response analysis (f(t) is arbitrary)
where f(t) is the dynamic force applied on the structure, t the time, and ω the circular frequency (Figure 8.1).
8.2 Review of Basic Equations
We begin with a review of vibration of a single-DOF system, consisting of a mass, a
spring, and a damper as shown in Figure 8.2. Then we will review basic equations
for a multi-DOF system, such as the discrete system resulting from the finite element
discretization.
225
226
Finite Element Modeling and Simulation with ANSYS Workbench
f(t)
FIGURE 8.1
A dynamic force applied to the structure.
k
m
c
FBD: ku
cu
m
f = f(t)
m—mass
k—stiffness
c—damping
f (t)—force
f(t)
x, u
FIGURE 8.2
A single DOF system with damping.
8.2.1 A Single DOF System
From the free-body diagram (FBD) and Newton’s law of motion (ma = f), we have:
= f (t) − k u − c u
mu
that is
+ c u + k u = f (t)
mu
(8.1)
= d 2u/dt 2 the acceleration.
where u is the displacement, u = du/dt the velocity, and u
Free Vibration (no applied force to the mass or f(t) = 0):
Free vibration occurs when a mass is moved away from its rest position due to initial
conditions.
Assuming zero damping (c = 0) in a free vibration, Equation 8.1 becomes:
+ k u = 0
mu
(8.2)
The physical meaning of this equation is: inertia force + elastic/stiffness force = 0.
Although there is no applied force, the mass can have nonzero displacement or experience
vibrations under the initial conditions. To solve for such nontrivial solutions, we assume:
u(t) = U sin ωt
where ω is the circular frequency of oscillation and U the amplitude. Substituting this into
Equation 8.2 yields:
227
Structural Vibration and Dynamics
− U ω 2 . m sin ω t + k U sin ω t = 0
that is
[−ω 2 m + k ] U = 0
For nontrivial solutions for U, we must have:
[−ω 2 m + k ] = 0
which yields
ω=
k
m
(8.3)
This is the circular natural frequency of the single DOF system (rad/s). The cyclic frequency (1/s = Hz) is ω/2π.
Equation 8.3 is a very important result in free vibration analysis, which says that
the natural frequency of a structure is proportional to the square-root of the stiffness
of the structure and inversely proportional to the square-root of the total mass of the
structure.
The typical response of the system in undamped free vibration is sketched in Figure 8.3.
For nonzero damping c, where
0 < c < cc = 2mω = 2 k m
(cc = critical damping )
(8.4)
we have the damped natural frequency:
ωd = ω 1 − ξ2
(8.5)
ξ = c/cc
(8.6)
where
is called the damping ratio.
u
u = U sin ω t
T = 1/f
FIGURE 8.3
Typical response in an undamped free vibration.
U
U
t
228
Finite Element Modeling and Simulation with ANSYS Workbench
u
t
FIGURE 8.4
Typical response of a free vibration with a nonzero damping c < cc .
For structural damping: 0 ≤ ξ < 0.15 (usually 1 ~ 5%)
ωd ≈ ω
(8.7)
That is, we can ignore damping in normal mode analysis.
The typical response of the system in damped free vibration is sketched in Figure 8.4.
We can see that damping has the effect of reducing the vibration of the system.
8.2.2 A Multi-DOF System
For a multi-DOF system, the equation of motion can be written as
+ Cu + Ku = f(t)
Mu
(8.8)
in which:
u—nodal displacement vector;
M—mass matrix;
C—damping matrix;
K—stiffness matrix;
f—forcing vector.
The physical meaning of Equation 8.8 is
Inertia forces + Damping forces + Elastic forces = Applied forces
We already know how to determine the stiffness matrix K for a structure, as discussed
in previous chapters. In vibration analysis, we also need to determine the mass matrix and
damping matrix for the structure.
8.2.2.1 Mass Matrices
There are two types of mass matrices: lumped mass matrices and consistent mass matrices. The
former is empirical and easier to determine, and the latter is analytical and more involved
in their computing.
We use a bar element to illustrate the lumped mass matrix (Figure 8.5).
For this bar element, the lumped mass matrix for the element is found to be
229
Structural Vibration and Dynamics
m1 =
ρAL
2
ρ, A, L
1
u1
2
m2 =
u2
ρAL
2
FIGURE 8.5
The lumped mass for a 1-D bar element.
ρAL
2
m=
0
0
ρAL
2
which is a diagonal matrix and thus is easier to compute.
In general, we apply the following element consistent mass matrix:
m=
∫ ρN N dV
T
(8.9)
V
where N is the same shape function matrix as used for the displacement field and V is the
volume of the element.
Equation 8.9 is obtained by considering the kinetic energy within an element:
1 T
1
(cf. mv 2 )
u mu
2
2
1
1
=
ρu 2 dV =
ρ(u )T u dV
2 V
2 V
1
=
ρ(Nu )T (Nu ) dV
2 V
1
= u T ρNT N dV u
2 V
Κ=
∫
∫
∫
∫
(8.10)
m
For the bar element (linear shape function), the consistent mass matrix is
m=
1 − ξ
1 − ξ
ξ
∫ ρN N dV = ∫ ρ
T
V
1/3
= ρAL
1/6
V
1
1/6 u
2
1/3 u
ξ AL dξ
(8.11)
which is a nondiagonal matrix.
Similar to the formation of the global stiffness matrix K, element mass matrices are
established in local coordinates first, then transformed to global coordinates, and finally
assembled together to form the global structure mass matrix M.
230
Finite Element Modeling and Simulation with ANSYS Workbench
v1
v2
θ1
ρ, A, L
θ2
FIGURE 8.6
The consistent mass for a 1-D simple beam element.
For a simple beam element (Figure 8.6), the consistent mass matrix can be found readily by
applying the four shape functions listed in Equation 3.6. We have:
m=
∫ ρN N dV
T
V
156
ρAL 22L
=
420 54
−13L
22L
4L2
13L
−3L2
54
13L
156
−22L
−13L
−3L2
−22L
4L2
v1
θ1
v2
θ2
(8.12)
8.2.2.2 Damping
There are two commonly used models for viscous damping: proportional damping (also
called Rayleigh damping) and modal damping.
In the proportional damping model, the damping matrix C is assumed to be proportional
to the stiffness and mass matrices in the following fashion:
C = αK + βM
(8.13)
where the constants α and β are found from the following two equations:
ξ1 =
αω1
β
β
αω 2
+
, ξ2 =
+
2
2ω1
2
2ω 2
(8.14)
with ω1, ω2, ξ1 and ξ2 (damping ratios) being specified by the user. The plots of the above
two equations are shown in Figure 8.7.
In the modal damping model, the viscous damping is incorporated in the modal equations.
The modal damping can be introduced as
2ξ1 ω1
0
Cφ =
0
0
2ξ 2 ω 2
2ξ n ω n
0
where ξi is the damping ratio at mode i of a n-DOF system.
(8.15)
231
Structural Vibration and Dynamics
Damping ratio
ξ=
β
αω
+
2
2ω
Design
spectrum
ξ2
Stiffness-proportional
dampling: ξ = αω , β = 0
2
Mass-proportional
dampling: ξ = β , α = 0
2ω
ξ1
0
0
ω1
ω2
Frequency
FIGURE 8.7
Two equations for determining the proportional damping coefficients. (R. D. Cook, Finite Element Modeling
for Stress Analysis, 1995, Hoboken, NJ, Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with
permission.)
8.3 Formulation for Modal Analysis
Modal analysis sets out to study the inherent vibration characteristics of a structure,
including:
• Natural frequencies
• Normal modes (shapes)
Let f(t) = 0 and C = 0 (ignore damping) in the dynamic Equation 8.8 and obtain:
+ Ku = 0
Mu
(8.16)
Assume that displacements vary harmonically with time, that is:
u(t) = u sin(ωt),
u (t) = ωu cos(ωt),
(t) = −ω 2 u sin(ωt)
u
where u is the vector of the amplitudes of the nodal displacements.
Substituting these into Equation 8.16 yields:
[K − ω 2M]u = 0
(8.17)
232
Finite Element Modeling and Simulation with ANSYS Workbench
This is a generalized eigenvalue problem (EVP). The trivial solution is u = 0 for any values of ω (not interesting). Nontrivial solutions (u ≠ 0) exist only if:
K − ω 2M = 0
(8.18)
This is an n-th order polynomial of ω2, from which we can find n solutions (roots) or
eigenvalues ωi (i = 1, 2, …, n). These are the natural frequencies (or characteristic frequencies) of the structure.
The smallest nonzero eigenvalue ω1 is called the fundamental frequency.
For each ωi, Equation 8.17 gives one solution or eigen vector:
[K − ω i 2M]ui = 0
ui (i = 1, 2, …, n) are the normal modes (or natural modes, mode shapes, and so on).
Properties of the Normal Modes:
Normal modes satisfy the following properties:
uTi Ku j = 0, uTi Mu j = 0, for i ≠ j
(8.19)
if ω i ≠ ω j . That is, modes are orthogonal (thus independent) to each other with respect to K
and M matrices.
Normal modes are usually normalized:
uTi Mui = 1, uTi Kui = ω i2
(8.20)
Notes:
• Magnitudes of displacements (modes) or stresses in normal mode analysis have
no physical meaning.
• For normal mode analysis, no support of the structure is necessary.
• ωi = 0 means there are rigid-body motions of the whole or a part of the structure.
This can be applied to check the FEA model (check to see if there are mechanisms
or free elements in the FEA models).
• Lower modes are more accurate than higher modes in the FEA calculations (due
to less spatial variations in the lower modes leading to that fewer elements/wavelengths are needed).
EXAMPLE 8.1
Consider the free vibration of a cantilever beam with one element as shown below.
y
v2
ρ, A, EI
1
L
θ2
2
x
233
Structural Vibration and Dynamics
We have the following equation for the free vibration (EVP):
v2 0
[K − ω 2M] =
θ 2 0
where
K=
EI
L3
12
−6L
−6L
ρAL 156
M=
2 ,
4L
420 −22L
−22L
4L2
The equation for determining the natural frequencies is
12 − 156λ
−6L + 22Lλ
−6L + 22Lλ
=0
4L2 − 4L2 λ
in which λ = ω 2 ρAL4 /420 EI .
Solving the EVP, we obtain:
EI
ω1 = 3.533
ρAL4
EI
ω 2 = 34.81
ρAL4
1
1
2
2
,
1
v2
= 1.38 ,
θ2 1
L
,
1
v2
= 7.62
θ2 2
L
The exact solutions of the first two natural frequencies for this problem are
EI
ω1 = 3.516
ρAL4
1
2
,
EI
ω 2 = 22.03
ρAL4
1
2
We can see that for the FEA solution with one beam element, mode 1 is calculated
much more accurately than mode 2. More elements are needed in order to compute
mode 2 more accurately. The first three mode shapes of the cantilever beam is shown in
the insert above.
8.3.1 Modal Equations
Use the normal modes (modal matrices), we can transform the coupled system of dynamic
equations to uncoupled system of equations or modal equations.
We have:
[K − ω i 2M]ui = 0,
i = 1, 2, ..., n
where the normal modes ui satisfy:
T
ui Ku j = 0,
for i ≠ j
T
ui Mu j = 0,
(8.21)
234
Finite Element Modeling and Simulation with ANSYS Workbench
and
T
ui Mui = 1,
for i = 1, 2, …, n
T
2
ui Kui = ω i ,
Form the modal matrix:
Φ( n × n ) = [ u1 u 2 u n ]
(8.22)
We can verify that:
ω12
0
ΦT KΦ = Ω =
0
0
ω 22
0
0
(Spectral matrix)
0
ω 2n
(8.23)
Φ MΦ = I.
T
Transformation for the displacement vector:
u = z1u1 + z2u 2 + + znu n = Φz
(8.24)
where
z1(t)
z (t)
2
z=
zn (t)
are called the principal coordinates.
Substitute Equation 8.24 into the dynamic Equation 8.8 and obtain:
MΦ
z + CΦz + KΦz = f(t)
Premultiply this result by ΦT, and apply Equation 8.23:
z + Cϕ z + Ωz = p(t)
(8.25)
where Cφ = αI + βΩ if proportional damping is applied, and p = Φ T f(t).
If we employ modal damping:
2ξ1 ω1
0
Cφ =
0
0
2ξ 2 ω 2
2ξ n ω n
0
235
Structural Vibration and Dynamics
where ξi is the damping ratio at mode i, Equation 8.25 becomes:
zi + 2ξ iω i z i + ω i2 zi = pi (t), i = 1, 2,…, n
(8.26)
Equations in Equation 8.25 with modal damping, or in Equation 8.26, are called modal equations. These equations are uncoupled, second-order differential equations, that are much
easier to solve than the original dynamic equation which is a coupled system.
To recover u from z, apply the transformation in Equation 8.24 again, once z is obtained
from Equation 8.26.
Notes:
• Only the first few modes may be needed in constructing the modal matrix Φ (i.e.,
Φ could be an n × m rectangular matrix with m < n). Thus, significant reduction in
the size of the system can be achieved.
• Modal equations are best suited for structural vibration problems in which
higher modes are not important (i.e., for structural vibrations, but not for structures under impact or shock loadings).
8.4 Formulation for Frequency Response Analysis
For frequency response analysis (also referred to as harmonic response analysis), the
applied dynamic load is a sine or cosine function. In this case, the equation of motion is
+ Cu + Ku =
Mu
F
sin
ω
t
Harmonic loading
(8.27)
8.4.1 Modal Method
In this approach, we apply the modal equations, that is
zi + 2ξ iω i z i + ω i2 zi = pi sin ωt i = 1, 2, …, m
(8.28)
These are uncoupled equations. The solutions for z are in the form:
zi (t) =
pi ω i2
(1 − ηi2 )2 + (2ξ i ηi )2
sin(ωt − θi )
where
2ξ i ηi
θi = arctan 1 − η2 , phase angle;
i
η
=
ω
ω
;
i
i
ci
ci
ξ i =
=
mping ratio
, dam
cc
2mω i
(8.29)
236
Finite Element Modeling and Simulation with ANSYS Workbench
The response of each mode Zi is similar to that of a single DOF system. Once the natural
coordinate vector z is known, we can recover the real displacement vector u from z using
Equation 8.24.
8.4.2 Direct Method
In this approach, we solve Equation 8.27 directly, that is, compute the inverse of the coefficient matrix, which is in general much more expensive than the modal method.
Using complex notation to represent the harmonic response, we have u = ue iω t and
Equation 8.27 becomes:
[K + iωC − ω 2M]u = F
(8.30)
Inverting the matrix [K + iωC − ω 2M], we can obtain the displacement amplitude
vector u. However, this equation is expensive to solve for large systems and the matrix
[K + iωC − ω 2M] can become ill-conditioned if ω is close to any natural frequency ωi of the
structure. Therefore, the direct method is only applied when the system of equations is
small and the frequency is away from any natural frequency of the structure.
8.5 Formulation for Transient Response Analysis
In transient response analysis (also referred to as dynamic response/time-history a­ nalysis),
we are interested in computing the responses of the structures under arbitrary time-­
dependent loading (Figure 8.8).
(a)
f(t)
t
(b)
u(t)
t
FIGURE 8.8
(a) A step type of loading; (b) structural response to the step loading.
237
Structural Vibration and Dynamics
u(t)
u1
uk uk + 1
u2
t0 t1 t2
tk tk + 1
t
FIGURE 8.9
Computing the responses by integration through time.
To compute the transient responses, integration through time is employed (Figure 8.9).
We write the equation of motion at instance tk, k = 0, 1, 2, 3, …, as
k + Cu k + Ku k = fk
Mu
(8.31)
Then, we introduce time increments: Δt = tk + 1 − tk, k = 0, 1, 2, 3, …, and integrate through
the time.
There are two categories of methods for transient analysis as described in the following
sections.
8.5.1 Direct Methods (Direct Integration Methods)
Central Difference Method:
Approximate the velocity and acceleration vectors by using finite difference:
1
(u k + 1 − u k −1 ),
2∆t
1
k =
(u k + 1 − 2u k + u k −1 )
u
(∆t)2
u k =
(8.32)
Dynamic equation becomes,
1
1
M
(u k + 1 − 2u k + u k −1 ) + C
(u k + 1 − u k −1 ) + Ku k = fk
2
2
(
∆
t
)
∆
t
which yields
Au k + 1 = F(t)
(8.33)
238
Finite Element Modeling and Simulation with ANSYS Workbench
where
1
1
A = (∆t)2 M + 2∆t C,
F(t) = fk − K − 2 M u k − 1 M − 1 C u k −1
(∆t)2
(∆t)2
2∆t
We compute u k+1 from u k and u k−1, which are known from the previous time step. The
solution procedure is repeated or marching from t0 , t1, … tk , tk + 1, … , until the specified maximum time is reached. This method is unstable if Δt is too large.
Newmark Method:
We use the following approximations:
(∆t)2
k + 2βu
k + 1 ], → (u
k + 1 = )
[(1 − 2β)u
2
k + γ u
k + 1 ]
≈ u k + ∆t[(1 − γ )u
u k + 1 ≈ u k + ∆tu k +
u k + 1
(8.34)
where β and γ are chosen constants. These lead to the following equation:
Au k + 1 = F(t)
(8.35)
where
γ
1
C+
M,
β∆t
β(∆t)2
k )
F(t) = f (fk + 1 , γ , β, ∆t , C, M, u k , u k , u
A=K+
This method is unconditionally stable if
2β ≥ γ ≥
1
2
For example, we can use γ = 1/2, β = 1/4, which gives the constant average acceleration
method.
Direct methods can be expensive, because of the need to compute A−1, repeatedly for
each time step if nonuniform time steps are used.
8.5.2 Modal Method
In this method, we first do the transformation of the dynamic equations using the modal
matrix before the time marching:
m
u=
∑ u z (t) = Φz,
i i
i =1
zi + 2ξ iω i z i + ω i zi = pi (t), i = 1, 2,… , m.
(8.36)
239
Structural Vibration and Dynamics
TABLE 8.1
Comparisons of the Methods
Direct Methods
Small models
More accurate (with small Δt)
Single loading
Shock loading
…
Modal Method
Large models
Higher modes ignored
Multiple loading
Periodic loading
…
Then, solve the uncoupled equations using an integration method. We can use, for example, 10% of the total modes (m = n/10). The advantages of the modal method are as follows:
•
•
•
•
Uncoupled system
Fewer equations
No inverse of matrices
More efficient for large problems
However, the modal method is less accurate if higher modes are important, which is the
case for structures under impact or shock loading. Table 8.1 summarizes the advantages
and disadvantages of the direct and modal methods for transient response analysis.
8.6 Modeling Examples
8.6.1 Modal Analysis
Take a bumper system as an example. Figure 8.10 shows a front bumper and the supporting
brackets in a car. The model is applied to study the dynamic responses of the bumper. Shell
elements were used for this study to obtain the natural frequencies and vibration modes.
Figure 8.11 shows the first mode of the bumper when it is constrained at the bracket locations.
FIGURE 8.10
FEA model of a front bumper and supporting brackets.
240
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 8.11
The first vibration mode of the bumper.
8.6.2 Frequency Response Analysis
Figure 8.12 shows the frequency response of the bumper system by the leftmost curve and
with the acceleration of the two brackets as the input. Several modifications of the bumper
design are also studied with the goal to increase the base natural frequency (e.g., from
below 30 Hz to above 35 Hz) and to reduce the magnitudes of the frequency responses.
The improved responses are shown by the other three curves in Figure 8.12.
8.6.3 Transient Response Analysis
One of the most interesting applications of the transient analysis with FEA is to conduct
crash analysis and virtual drop tests of various products. Figure 8.13 is an example of
crash analysis of a car using the FEA, showing that the bumper system absorbs the
shock energy to help reduce the damage to the car.
8.6.4 Cautions in Dynamic Analysis
• Symmetry model should not be used in the dynamic analysis (normal modes,
etc.) because symmetric structures can have nonsymmetric modes. However,
Total displacement (mm)
5.0
Proposed I model (current design)
Proposed III model
Proposed III model
Reference model
4.0
3.0
2.0
1.0
0.0
0.0
10.0
30.0
20.0
Frequency (Hz)
FIGURE 8.12
Frequency response of the bumper from 0 to 50 Hz.
40.0
50.0
241
Structural Vibration and Dynamics
FIGURE 8.13
Car crash analysis using the FEA. (From LS-Dyna website, http://www.dynaexamples.com/)
TABLE 8.2
Units in Dynamic Analysis
t (time)
L (length)
m (mass)
a (accel.)
f (force)
ρ (density)
Choice I
Choice II
s
m
kg
m/s2
N
kg/m3
S
mm
Mg
mm/s2
N
Mg/mm3
symmetry can still be applied in creating the FEA model of a symmetric
structure.
• Mechanism or rigid body motion means ω = 0. Can use this to check FEA models
to see if they are properly connected and/or supported.
• Input for FEA: Loading F(t) or F(ω) can be very complex and data can be enormous
in real engineering applications (e.g., the load data for a car) and thus they often
need to be filtered first before being used as input for FEA.
• Selecting a proper unit system is very important in vibration or dynamic analysis.
Two choices of the units are listed in Table 8.2. Make sure they are consistent in
the FEA models.
8.7 Case Studies with ANSYS Workbench
Problem Description: Musical instruments such as acoustic guitars create sound by
means of vibration and resonance. The body of an acoustic guitar acts as a resonating
chamber when the strings are set into oscillation at their natural frequencies. The following figure gives the dimensions of a simplified acoustic guitar model. The guitar has
a wall thickness of 3 mm, and is made of Douglas fir wood (E = 13.1 GPa, Poison’s ratio
ν = 0.3, density = 470 kg/m­3). Assuming the back surface of the guitar is fixed, find the
first 10 natural frequencies and plot the first five vibration modes of the guitar. Suppose
a harmonic pressure loading of magnitude 1 MPa is applied to a side wall of the guitar.
Plot the frequency response of the z displacement (along the surface normal direction) of
the front surface.
242
Finite Element Modeling and Simulation with ANSYS Workbench
Side wall
Front face
H9
H10
C
H12
D
E
G
F
V2
B
A
H11
H15
H14
H13
V4
V3
V1
V5
H16
H17
K D18
V6
H
V7
I V8
J
Material:
Douglas fir (E = 13.1 GPa, v = 0.3, density
= 470 kg/m3)
Boundary conditions:
A fixed back face; harmonic pressure of
1 MPa applied to a side wall.
Construction point coordinates:
Point x (mm) y (mm)
A
0
0
B
0
10
C
30
70
D
60
80
E
100 70
F
140 50
G
200 60
H
220 50
I
240 10
J
240 0
K
170 0
The guitar profile is a spline that goes
through points A through J.
A circular hole centered at K has a
diameter of 45 mm.
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Guitar.wbpj.”
Step 2: Create a Modal Analysis System
Drag the Modal icon from the Analysis Systems Toolbox window and drop it inside
the highlighted green rectangle in the Project Schematic window to create a
standalone modal analysis system.
Step 3: Add a New Material
Double-click on the Engineering Data cell to add a new material. In the following
Engineering Data interface which replaces the Project Schematic, type “Wood” as
the name for the new material, and double-click Isotropic Elasticity under Linear
Structural Vibration and Dynamics
Elastic in the leftmost Toolbox window. Enter “13.1E9” for Young’s Modulus and
“0.3” for Poisson’s Ratio in the Properties window. Double-click Density under
Physical Properties. Enter “470” for Density in the Properties window. Click the
Return to Project button to go back to Project Schematic.
Step 4: Launch the DesignModeler Program
Ensure Surface Bodies is checked in the Properties of Schematic A3: Geometry window (select Properties from the View drop-down menu to enable display of this
window). Choose 3D as the Analysis Type in this Properties window. Doubleclick the Geometry cell to launch DesignModeler, and select “Millimeter” in the
Units pop-up window.
Step 5: Create a Profile Sketch
Click on the Sketching tab. Select the Draw toolbox and then Construction Point.
Draw 10 construction points A through J, as shown below. Draw a spline passing through points A through J; right-click at the last construction point and
choose Open End from the context menu to finish the spline creation.
243
244
Finite Element Modeling and Simulation with ANSYS Workbench
C
D
E
F
B
A
G
H
I
J
Step 6: Create a Replicate Curve
Select the Modify toolbox and then Replicate. Click on the spline from the Graphics
window. Right-click anywhere in the Graphics to show the context menu. Select
End/Use Plane Origin as Handle as shown below. A replicate spline will appear
in the Graphics window.
Structural Vibration and Dynamics
Next, right-click anywhere in the Graphics, and select Flip Vertical in the context
menu. A vertically flipped spline will appear.
Click on the origin point in the Graphics to place the flipped spline, and press
Esc to end the operation. A closed-loop curve is now formed as shown below.
245
246
Finite Element Modeling and Simulation with ANSYS Workbench
Step 7: Create an Extruded Body
Switch to the Modeling tab and click on the Extrude feature. The default Base Object
is set as Sketch1 in the Details of Extrude1. Change the extrusion depth to 50 mm
in the field of FD1, Depth and click Generate. A solid body is created as shown
below.
Step 8: Create an Extruded Cut on the Front Face
Create a new plane by selecting New Plane from the Create drop-down menu.
Structural Vibration and Dynamics
A new plane named Plane4 is now added to the Tree Outline. In the Details of
Plane4, set the Type to From Face. Click the front face of the guitar from the
Graphics window, and apply it to the Base Face selection in the Details of Plane4.
Click Generate.
To create a new sketch under Plane4 in the Tree Outline, click on the New Sketch
icon.
New
sketch
247
248
Finite Element Modeling and Simulation with ANSYS Workbench
Switch to the Sketching tab for Sketch2. In the sketch, draw a horizontal line by connecting points A and B as shown below. Then draw a circle of diameter 45 mm
centered at point C, located 170 mm to the left of point A along line AB.
A
C
B
Next, choose Trim under the Modify tab, and click on line AB in the Graphics window. The sketch line AB will disappear. Click Generate.
249
Structural Vibration and Dynamics
Switch to the Modeling tab, and click on the Extrude feature. The default Base Object
is set as Sketch2 in the Details of Extrude2. Set the Operation to Cut Material. Enter
an extrusion depth of 10 mm in the field of FD1, Depth and click Generate. An
extruded cut feature is now added to the front face as shown below.
Step 9: Create a Surface Body
Select Surface from Faces from the Concept drop-down menu. In the Graphics window, Ctrl-click to select four faces, that is, the front, back, top, and bottom faces
that enclose the solid body as shown below.
Top face
Back face
Front face
Bottom face
250
Finite Element Modeling and Simulation with ANSYS Workbench
Click Apply next to Faces in the Details of SurfFromFaces1. Then click Generate. A
surface body will be generated in the Tree Outline under 2 Parts, 2 Bodies.
Right-click on Solid under 2 Parts, 2 Bodies in the Tree Outline. In the context menu,
select Suppress Body.
Structural Vibration and Dynamics
Click on Surface Body under 2 Parts, 2 Bodies in the Tree Outline. Change the
Thickness to 3 mm in the Details of Surface Body. Save and exit the DesignModeler.
Step 10: Launch the Modal–Mechanical Program
Double-click on the Model cell to launch the Modal–Mechanical program. Click
on the Surface Body under Geometry in the Outline tree. In the Details of “Surface
Body,” click to the right of the Material Assignment field and select Wood from
the drop-down menu.
251
252
Finite Element Modeling and Simulation with ANSYS Workbench
Step 11: Generate Mesh
Right click on Mesh in the Project Outline. Select Insert and then Sizing from the
context menu. In the Details of “Face Sizing,” enter “5e-4 m” for the Element Size.
Click on the front, back, top, and bottom faces of the guitar in the Graphics window and apply the four faces to the Geometry selection.
Right-click on Mesh in the Outline, and select Generate Mesh from the context menu.
Structural Vibration and Dynamics
Step 12: Set Up Modal Analysis and Apply Boundary Conditions
Click on Analysis Settings under Modal in the Outline tree. Change the Max Modes
to Find to 10 in the Details of “Analysis Settings.”
Right-click on Modal(A5). Choose Insert and then Fixed Support from the context
menu. Apply the back face to the Geometry selection.
253
254
Finite Element Modeling and Simulation with ANSYS Workbench
Step 13: Retrieve Results from Modal Analysis
Insert Total Deformation by right-clicking on Solution (A6) in the Outline. In the
Details of “Total Deformation,” set Mode to 1. Insert another Total Deformation
item. In the Details of “Total Deformation 2,” set Mode to 2. Repeat this step three
more times. Set Mode to 3, 4, and 5, respectively, for each new insertion. Rightclick on Solution (A6) in the Outline and Solve
Click on Total Deformation in the Outline to review results. The results below show
the first natural frequency of 1036.8 Hz and the corresponding mode shape.
Click on
tabular
data and
graph.
Structural Vibration and Dynamics
Click on Tabular Data and Graph on the right edge of the Graphics window, and
then click on the push pin labeled AutoHide to display the Tabular data and the
Graph as shown below.
The Tabular data gives the first 10 natural frequencies of the guitar under the
fixed bottom boundary condition. The Play/Stop control interface in the Graph
window allows animation of mode shapes. Click on each different Total
Deformation item in the Outline to review results, for example, the following figure shows the fourth mode shape, and then exit the Modal–Mechanical
program.
255
256
Finite Element Modeling and Simulation with ANSYS Workbench
Step 14: Create a Harmonic Response Analysis System
Drag the Harmonic Response icon from the Analysis Systems Toolbox window and
drop it onto the Solution cell of the highlighted Modal system in the Project
Schematic.
This creates a Harmonic Response system that shares data with Modal system as
shown below.
Step 15: Set Up Harmonic Response Analysis and Assign Loads
Double-click on the Setup cell of the Harmonic Response system to launch the
Multiple Systems–Mechanical program. In the program, select Analysis Settings
Structural Vibration and Dynamics
from the Outline. Set the Range Minimum to 1000 Hz, Range Maximum to 4000 Hz,
and Solution Intervals to 300.
Right-click on Harmonic Response (B5). Choose Insert and then Pressure from the
context menu. In the Details of “Pressure,” set magnitude as 1 MPa, and apply the
top face to the Geometry selection.
Step 16: Retrieve Results from Harmonic Response
Right-click on Solution (B6). Choose Insert and Frequency Response and then
Deformation from the context menu. In the Details of “Frequency Response,” set
257
258
Finite Element Modeling and Simulation with ANSYS Workbench
the Orientation of the directional deformation to Z-Axis. Click on the front face
and apply it to the Geometry selection.
Click on
the front
face.
Right-click on Solution (B6) and select Solve. After solution is done, click on Frequency
Response in the Outline to review the harmonic response of the guitar.
Modeling tips: Note that modal analysis can be run as constrained, unconstrained, or
partially constrained. Also note that symmetric structure may have asymmetric modes,
and thus, it is not recommended to take advantage of symmetry to simplify models for
modal analysis. In the following, we are going to show a result comparison between an
unconstrained model and the fixed model. For the unconstrained model, the first six mode
shapes obtained from simulation are rigid body modes that allow the structure to move
freely. They are not considered as structural modes. The comparison indicates that a free
floating guitar has a different set of natural frequencies and mode shapes from that of a
fixed guitar. In general, constraint conditions have an effect on the vibration characteristics of a structure and should be considered carefully when setting up a model.
259
Structural Vibration and Dynamics
A: Model
Total deformation
Type: Total deformation
Frequency: 1018.6 Hz
Unit: m
A: Model
Total deformation 2
Type: Total deformation
Frequency: 1278.3 Hz
Unit: m
9.2507 max
8.2478
7.2449
6.2419
5.239
4.236
3.2331
2.2302
1.2272
0.0063643 min
8.9597 max
7.9649
6.9701
5.9753
4.9805
3.9856
2.9908
1.996
1.0012
0.0063643 min
1018.6 Hz
A: Model
Total deformation 3
Type: Total deformation
Frequency: 1573.1 Hz
Unit: m
1278.3 Hz
A: Model
Total deformation 4
Type: Total deformation
Frequency: 1798.8 Hz
Unit: m
15.406 max
13.702
11.997
10.293
8.5889
6.8847
5.1805
3.4763
1.7721
0.067943 min
11.861 max
10.544
9.2269
7.9098
6.5926
5.2754
3.9583
2.6411
1.324
0.0068332 min
1573.1 Hz
1798.8 Hz
The first four structural modes of the unconstrained guitar model.
A: Model
Total deformation 2
Type: Total deformation
Frequency: 1683.8 Hz
Unit: m
A: Model
Total deformation
Type: Total deformation
Frequency: 1036.8 Hz
Unit: m
12.569 max
11.172
9.7759
8.3794
6.9828
5.5862
4.1897
2.7931
1.3966
0 min
17.301 max
15.379
13.456
11.534
9.6117
7.6894
5.767
3.8447
1.9223
0 min
1036.8 Hz
A: Model
Total deformation 3
Type: Total deformation
Frequency: 2157.1 Hz
Unit: m
1683.8 Hz
A: Model
Total deformation 4
Type: Total deformation
Frequency: 2275.5 Hz
Unit: m
20.013 max
17.789
15.566
13.342
11.118
8.8946
6.671
4.4473
2.2237
0 min
13.125 max
11.666
10.208
8.7497
7.2914
5.8332
4.3749
2.9166
1.4583
0 min
2157.1 Hz
2275.5 Hz
The first four structural modes of the fixed guitar model.
260
Finite Element Modeling and Simulation with ANSYS Workbench
8.8 Summary
In this chapter, we first reviewed the equations of motion for both single DOF and multiple DOF systems and discussed how to compute the mass and damping matrices in the
FE formulations. Then, we discussed the methods for solving normal modes, harmonic
responses, and transient responses for structural vibration and dynamic problems. The
advantages and disadvantages of the direct method and modal method are discussed.
Several examples of vibration and dynamic analyses are also discussed to show the applications of the FEA in vibration and dynamic analyses. The vibration modes of a guitar are
analyzed using ANSYS Workbench.
8.9 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1. Understand the FE formulations for vibration and dynamic analysis of structures
2. Know how to compute the mass and damping matrices in the FE equations of
motion
3. Solve for the natural frequencies and modes of structures using the FEA
4. Compute the harmonic response of structures using the FEA
5. Compute the transient response of structures using the FEA
6. Perform vibration or dynamic analysis of structures using ANSYS Workbench
PROBLEMS
8.1 For the cantilever beam studied in Example 8.1, apply more elements and
investigate the FEA solutions for the first 10 natural frequencies and normal
modes of the beam structure. Assume the beam is made of structural steel
with Young’s modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and mass density
ρ = 7850 kg/m3, and has a length of 100 cm with a 10 cm × 10 cm square cross
section.
8.2 For a square plate with edge length = 1 m, thickness = 0.005 m, Young’s modulus E = 70 GPa, Poisson’s ratio ν = 0.3, and mass density ρ = 2800 kg/m3, find the
first five natural frequencies and normal modes using the FEA when the plate is
a. Clamped at the four edges
b. Simply supported at the four edges
c. Free at the four edges (not supported)
Discuss the effects of the support on the natural frequencies of the plate.
261
Structural Vibration and Dynamics
8.3 A fuel tank, with a total length = 5 m, diameter = 1 m, and thickness = 0.01 m,
is shown below. Using the FEA, find the first six natural frequencies and corresponding normal modes, when:
a. The tank is not supported at all.
b. The tank is constrained along the circumferences in the radial directions at
the two locations 1 m away from the two ends.
Carefully build your FE mesh (using shell elements) so that the symmetry of the
tank is reserved and the boundary conditions in part (b) can be applied readily.
Assume the tank is made of steel with Young’s modulus E = 200 GPa, Poisson’s
ratio ν = 0.3, and mass density ρ = 7850 kg/m3.
8.4 For the rotating part sketched below, assume that it is made of steel with
Young’s modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and mass density
ρ = 7850 kg/m3. Ignore the gravitational force. Compute the first five natural
frequencies and normal modes in the part using the FEA when the part is
a. Fixed on the inner cylindrical surface of the hole
b. Fixed on the bottom surface
Compare the results between the two models.
z
ω
12 mm
8 mm
30 mm
16 mm
50 mm
30 mm
5 mm
262
Finite Element Modeling and Simulation with ANSYS Workbench
8.5 The following figure shows a simple plate made of structural steel with Young’s
modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and density ρ = 7850 kg/m3. The
plate has a thickness of 5 mm and is fixed on the left side. Suppose a harmonic
force of magnitude 100 N is applied to the right edge along the plate’s surface
normal direction. Use a solid model and a shell model to compute:
a. The first five natural frequencies and the corresponding normal modes
b. The frequency response of the z directional deformation (along the surface normal direction of the plate) of the plate surface under the given harmonic load
Compare the results between the two models.
All dimensions are in centimeters.
5.000
R1.000
R5.000
10.000
20.000
10.000
40.000
8.6 The following figure shows a simple bracket made of structural steel with Young’s
modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and density ρ = 7850 kg/m3. The
bracket has a front face profile shown below and an extrusion depth of 12 cm.
A through-hole of diameter 1.5 cm is centered on the bracket’s flat ends, one on
each side. Suppose that the bracket is fixed on the bottom, and is subjected to a
harmonic pressure load of magnitude 1000 N/cm2 on the top face. Use a solid
model to compute:
a. The first five natural frequencies and the corresponding normal modes
b. The frequency response of the z directional deformation (along the front face
normal direction) of the front face of the bracket under the given harmonic load
Load
All dimensions are in
centimeters.
10.000
1.500
1.500
R3.000
1.500
8.000
1.500
12.000
8.000
1.500
263
Structural Vibration and Dynamics
8.7 The following figure shows an L-shaped bracket made of structural steel with
Young’s modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and density ρ = 7850 kg/m3.
The 4 cm thick L-shape with a cross-section profile shown in the figure is rigidly connected to a rectangular base of size 3 cm × 8 cm × 1 cm. Suppose that
the bracket is fixed on the right end at the base. A harmonic pressure of magnitude 100 N/cm2 is applied on the protruding surface of the L-shape as shown
below. Using the FEA, compute:
a. The first five natural frequencies and the corresponding normal modes
b. The frequency response of the y directional deformation (along the load
direction) of the bracket’s protruding top face under the given harmonic load
All dimensions are in
centimeters.
Load
1.000
R1.000
1.000
3.000
8.000
8.8 The following two-bar assembly consists of four components made of structural
steel with Young’s modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and density
ρ = 7850 kg/m3. As shown in the figure below, the two bars are identical in size.
Each bar has a diameter of 2 cm and a length of 16 cm. The two identical end
blocks each has a thickness of 2 cm and a fillet radius of 1 cm at the round corners. The distance of the gap in between the two blocks is 8 cm. Suppose that the
two blocks are fixed on the bottom. The bars are applied a harmonic pressure
load of magnitude 100 N/cm2 as shown below. Using the FEA, compute:
a. The first five natural frequencies and normal modes of the assembly
b. The frequency response of the x directional deformation (along the load
direction) of the cylindrical bar surface under the given harmonic load
All dimensions are in centimeters.
Use bonded contact for all contact regions.
2.000 5.000
Load
R1.000
Load
10.000
6.000
264
Finite Element Modeling and Simulation with ANSYS Workbench
8.9 The following clevis assembly consists of a yoke, a pin, and a u-shape. The
assembly components are made of structural steel with Young’s modulus
E = 200 GPa, Poisson’s ratio ν = 0.3, and density ρ = 7850 kg/m3. The profile
sketches of the yoke and the u-shape are shown below. Both components have
an extrusion depth of 5 cm. The pin has a diameter of 2 cm and a length of
16 cm, and is centered 2.5 cm away from both the front and the side faces of
the u-shape. Assume that the yoke is fixed on the left end, and the u-shape is
applied a harmonic pressure load of magnitude 100 N/cm2 on the right end.
Using the FEA, compute:
a. The first five natural frequencies and normal modes of the assembly
b. The frequency response of the x directional deformation (along the load
direction) of the cylindrical pin surface under the given harmonic load
Use no separation
for all contacts.
Load
A clevis assembly
The yoke and the u-shape
12.000
All dimensions are
in centimeters.
1.500
2.000
R1.500
1.500
3.000
R1.500
16.000
2.000
3.000
8.000
D2.000
2.500
2.000
2.500
265
Structural Vibration and Dynamics
8.10 The following rotor assembly consists of a rotor, a shaft, and a base. The assembly components are made of structural steel with Young’s modulus E = 200 GPa,
Poisson’s ratio ν = 0.3, and density ρ = 7850 kg/m3. The base of the assembly is
fixed on the bottom. Suppose a harmonic pressure load of magnitude 100 N/
cm2 is applied to the side surface of the base as shown below. Using the FEA,
compute:
a. The first five natural frequencies and normal modes of the assembly
b. The frequency response of the y directional deformation (in the vertical
direction) of the cylindrical shaft surface under the given harmonic load
1.000
Use bonded contact for
all contact regions.
3.000
2.000
20.000
2.000 10.000
6.000
28.000
Load
3.000
R3.000
40.000
A rotor assembly
Assembly front view
(unit: cm)
The base
The rotor
9
Thermal Analysis
9.1 Introduction
Many engineering problems are thermal problems in nature. Devices such as appliances,
advanced electronics, engines, and heating, ventilation, and air conditioning systems need
to be evaluated for their thermal performance during the design process. In this chapter, we will discuss thermal analysis using the FEA. The objective of thermal analysis is
to understand response and behavior of a structure with thermal loading. The resulting
temperature distribution, heat flux distribution, and structural response under different
thermal loading conditions constitute important knowledge in assuring design success
of thermal engineering products. Both steady-state and transient thermal problems are
introduced and a heat sink model is analyzed using ANSYS® Workbench. Thermal stress
analysis, which is to find the structural response due to change of temperatures, is also
included and discussed.
9.2 Review of Basic Equations
To study heat transfer within an object or between objects, one may conduct thermal analysis, from which thermal quantities such as the temperature, thermal gradient, and heat
flux distributions can be determined. There are two types of thermal analyses: steady-state
thermal and transient thermal. Steady-state thermal analysis aims to find the temperature or
heat flux distribution in structures when a thermal equilibrium is reached, and transient
thermal analysis sets out to determine the time history of how the temperature profile and
other thermal quantities change with time. In addition, thermal expansion or contraction
of engineering materials often leads to thermal stress in structures, which can be examined by conducting thermal stress analysis. The basic equations for thermal and thermal
stress analyses are given as follows.
9.2.1 Thermal Analysis
For temperature field in a 1-D space, such as a bar (Figure 9.1), we have the following
Fourier heat conduction equation:
fx = −k
∂T
∂x
(9.1)
267
268
Finite Element Modeling and Simulation with ANSYS Workbench
T(x, t)
x
FIGURE 9.1
The temperature field T(x, t) in a 1-D bar model.
where,
fx = heat flux per unit area
k = thermal conductivity
T = T(x, t) = temperature field
For 3-D case, we have:
fx
∂T ∂x
f y = − K ∂T ∂y
f
∂T ∂z
z
(9.2)
where, fx, f y, fz = heat flux in the x, y, and z direction, respectively. In the case of isotropic
materials, the conductivity matrix is:
k
K = 0
0
0
k
0
0
0
k
(9.3)
The equation of heat flow is given by
∂f
∂f y ∂ f z
∂T
− x +
+
+ qv = cρ
∂y
∂z
∂t
∂x
(9.4)
in which,
qv = rate of internal heat generation per unit volume
c = specific heat
ρ = mass density
For steady-state case (∂T/∂t = 0) and isotropic materials, we can obtain:
k∇ 2T = − qv
(9.5)
This is a Poisson equation, which needs to be solved under given boundary conditions.
Boundary conditions for steady-state heat conduction problems are (Figure 9.2):
T = T , on ST
Q ≡ −k
∂T
= Q , on Sq
∂n
(9.6)
(9.7)
Note that at any point on the boundary S = ST ∪ Sq , only one type of BCs can be specified.
269
Thermal Analysis
n
Sq
y
ST
x
FIGURE 9.2
Boundary conditions for heat conduction problems.
9.2.1.1 Finite Element Formulation for Heat Conduction
For heat conduction problems, we can establish the following FE equation:
KT T = q
(9.8)
where,
KT = conductivity matrix
T = vector of nodal temperature
q = vector of thermal loads
The element conductivity matrix is given by
kT =
∫ B KBdV
T
(9.9)
V
This is obtained in a similar way as for the structural analysis, that is, by starting with
the interpolation T = NTe for the temperature field (with N being the shape function matrix
and Te the nodal temperature). Note that there is only one DOF at each node for the thermal
problems.
For transient (unsteady state) heat conduction problems, we have:
∂T
≠0
∂t
In this case, we need to apply finite difference schemes (use time steps and integrate in
time), as in the transient structural analysis, to obtain the transient temperature fields.
9.2.2 Thermal Stress Analysis
To determine the thermal stresses due to temperature changes in structures, we can proceed to
• Solve Equation 9.8 first to obtain the temperature (change) fields.
• Apply the temperature change ΔT as initial strains (or initial stresses) to the structure to compute the thermal stresses due to the temperature change.
270
Finite Element Modeling and Simulation with ANSYS Workbench
9.2.2.1 1-D Case
To understand the stress–strain relations in cases where solids undergo temperature
changes, we first examine the 1-D case (Figure 9.3). We have for the thermal strain (or
initial strain):
εo = αΔT
(9.10)
in which
α = the coefficient of thermal expansion
ΔT = T2 − T1 = change of temperature
Total strain is given by
ε = εe + εo
(9.11)
with εe being the elastic strain due to mechanical load.
That is, the total strain can be written as
ε = E−1σ + αΔT
(9.12)
σ = E(ε – εo)
(9.13)
Or, inversely, the stress is given by
EXAMPLE 9.1
Consider the bar under thermal load ΔT as shown in Figure 9.3.
(a) If no constraint on the right-hand side, that is, the bar is free to expand to the
right, then we have:
ε = εo,
εe = 0, σ = 0
from Equation 9.13. That is, there is no thermal stress in this case.
(b) If there is a constraint on the right-hand side, that is, the bar cannot expand to
the right, then we have:
ε = 0, εe = −εo = −αΔT,
σ = −EαΔT
from Equations 9.11 and 9.13. Thus, thermal stress exists.
From this simple example, we see that the way in which the structure is constrained
has a critical role in inducing the thermal stresses.
At temperature T1
At temperature T2
εo
FIGURE 9.3
Expansion of a bar due to increase in temperature.
271
Thermal Analysis
9.2.2.2 2-D Cases
For plane stress, we have:
εx
α∆T
ε o = ε y = α∆T
γ
xy o 0
(9.14)
εx
(1 + ν)α∆T
ε o = ε y = (1 + ν)α∆T
γ
0
xy o
(9.15)
εx
α∆T
ε
α∆T
y
α∆T
ε z
εo = =
0
γ xy
0
γ yz
γ zx o 0
(9.16)
For plane strain, we have:
in which, ν is Poisson’s ratio.
9.2.2.3 3-D Case
Observation: Temperature changes do not yield shear strains.
In both 2-D and 3-D cases, the total strain can be given by the following vector equation:
ε = εe + εo
(9.17)
And the stress–strain relation is given by
σ = Eεe = E(ε − εo)
(9.18)
9.2.2.4 Notes on FEA for Thermal Stress Analysis
Need to specify α for the structure and ΔT on the related elements (which experience the
temperature change).
• Note that for linear thermoelasticity, same temperature change will yield same
stresses, even if the structure is at two different temperature levels.
• Differences in the temperatures during the manufacturing and working environment are the main cause of thermal (residual) stresses.
A more comprehensive review of thermal problems, their governing equations and
boundary conditions can be found in the references, such as References [13,14].
272
Finite Element Modeling and Simulation with ANSYS Workbench
9.3 Modeling of Thermal Problems
Heat transfers in three ways through conduction, convection, and radiation. In FEM,
conduction is modeled by solving the resulting heat balance equations for the nodal
temperatures under specified thermal boundary conditions. Convection is modeled as a
surface load with a user-specified heat transfer coefficient and a given bulk temperature
of the surrounding fluid. Radiation effects, which are nonlinear, are typically modeled
by using the radiation link elements or surface effect elements with the radiation option.
Material properties such as density, thermal conductivity, and specific heat are needed
as input parameters for transient thermal analysis, while steady-state thermal analysis
needs only thermal conductivity as the material input. For thermal stress analysis, material input parameters include Young’s modulus, Poisson’s ratio, and thermal expansion
coefficient. In the following, modeling of thermal problems is briefly illustrated with the
aid of two examples.
9.3.1 Thermal Analysis
First, we use a heat sink model taken from Reference [14] for thermal analysis. A heat sink
is a device commonly used to dissipate heat from a CPU in a computer. In this heat sink
model, a given temperature field (T = 120) is specified on the bottom surface and a heat flux
condition (Q ≡ − k ∂ T/ ∂ n = −0.2 ) is specified on all the other surfaces. An FE mesh with a
total node of 127,149 is created as shown in Figure 9.4. Using the steady-state thermal analysis system in ANSYS, the computed temperature distribution on the heat sink is calculated
as shown in Figure 9.5. The cooling effect of the heat sink is most evident.
9.3.2 Thermal Stress Analysis
Next, we study the thermal stresses in structures due to temperature changes. For this
purpose, we employ the same model of a plate with a center hole (Figure 9.6) as used in
Chapters 4 and 5 to show the relation between the thermal stresses and constraints. We
assume that the plate is made of steel with Young’s modulus E = 200 GPa, Poisson’s ratio
ν = 0.3, and thermal expansion coefficient α = 12 × 10−6/°C. The plate is applied with a uniform temperature increase of 100°C.
FIGURE 9.4
A heat-sink model used for heat conduction analysis.
273
Thermal Analysis
MN
106.657
108.14
109.622
111.105
112.587
114.07
115.552
117.035
118.517
120
FIGURE 9.5
Computed temperature distribution in the heat-sink.
FIGURE 9.6
A square plate with a center hole and under a uniform temperature load.
Figure 9.7 shows the computed thermal stresses in the plate under two different types
of constraints. When the plate is constrained (roller support) at the left side only, the
plate expands uniformly in both the x and y directions, which causes no thermal stresses
(Figure 9.7a, note that the numbers, ranging from 10−6 to 10−3, are actually machine zeros).
However, when the plate is constrained at both the left and right sides, the plate can expand
only in the y direction and significant thermal stresses are induced (Figure 9.7b), especially
near the edge of the hole.
In many cases, the changes of the material properties of a structure should be considered as well when the temperature changes are significant, especially when the structure
is exposed to high temperatures such as in an aircraft engine. Cyclic temperature fields
can also cause thermal fatigue of structures and lead to failures. All these phenomena can
be modeled with the FEA and interested readers can consult with the documents of the
FEA software at hand.
274
Finite Element Modeling and Simulation with ANSYS Workbench
(a)
.681E-06
.584E-04
.156E-03
.258E-03
.352E-03
.456E-04
.147E-03
.245E-03
.343E-03
.441E-03
(b)
.259E+08
.169E+09
.313E+09
.456E+09
.600E+09
.576E+08
.241E+09
.354E+09
.528E+09
.671E+09
FIGURE 9.7
Thermal (von Mises) stresses in the plate: (a) When the plate is constrained at the left side only (thermal
stresses = 0); (b) when the plate is constrained at both the left and right sides.
9.4 Case Studies with ANSYS Workbench
Problem Description: Heat sinks are commonly used to enhance heat dissipation from
electronic devices. In the case study, we conduct thermal analysis of a heat sink made
of aluminum with thermal conductivity k = 170 W/(m K), density ρ = 2800 kg/m3, specific heat c = 870 J/(kg K), Young’s modulus E = 70 GPa, Poisson’s ratio ν = 0.3, and thermal expansion coefficient α = 22 × 10−6/°C. A fan forces air over all surfaces of the heat
sink except for the base, where a heat flux q′ is prescribed. The surrounding air is 28°C
with a heat transfer coefficient of h = 30 W/(m2°C). (Part A): Study the steady-state thermal response of the heat sink with an initial temperature of 28°C and a constant heat
flux input of q′ = 1000 W/m2. (Part B): Suppose the heat flux is a square wave function
with period of 90 s and magnitudes transitioning between 0 and 1000 W/m2. Study the
transient thermal response of the heat sink in 180 s by using the steady-state solution as
the initial condition. (Part C): Suppose the base of the heat sink is fixed. Study the thermal
stress response of the heat sink by using the steady-state solution as the temperature load.
275
Thermal Analysis
Material: Aluminum
k = 170 W/(m · K)
ρ = 2800 kg/m3; c = 870 J/(kg · K)
E = 70 GPa; v = 0.3
α = 22 × 10–6/°C
Base of the
heat sink
3.000
1.000
1.000
1.000
Boundary conditions:
Air temperature of 28°C; h = 30 W/(m2 · °C).
Steady state: q′ = 1000 W/m2 on the base.
8.000
2.000
2.000
Transient: Square wave heat flux on the base.
Initial conditions:
3.000
22.000
All dimensions are in millimeters.
Steady state: Uniform temperature of 28°C.
Transient: Steady-state temperature results.
Part A: Steady-State Thermal Analysis
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as HeatSink.wbpj.
Step 2: Create a Steady-State Thermal Analysis System
Drag the Steady-State Thermal icon from the Analysis Systems Toolbox window and drop
it inside the highlighted green rectangle in the Project Schematic window.
Step 3: Add a New Material
Double-click on the Engineering Data cell to add a new material. In the following
Engineering Data interface which replaces the Project Schematic, type “Aluminum”
276
Finite Element Modeling and Simulation with ANSYS Workbench
as the name for the new material, and double-click Isotropic Thermal Conductivity
under Thermal in the Toolbox window. Change the Unit to “Wm^-1K^-1” and enter
“170” for Isotropic Thermal Conductivity in the Properties window. Click the Return
to Project button to go back to Project Schematic.
Step 4: Launch the DesignModeler Program
Double-click the Geometry cell to launch DesignModeler, and select “Millimeter” in the
Units pop-up window.
Thermal Analysis
277
Step 5: Create a Profile Sketch
Click on the Sketching tab. Select the Draw toolbox and draw a 2D profile as shown
below.
Step 6: Create an Extruded Body
Switch to the Modeling tab and click on the Extrude feature. The default Base Object is set
as Sketch1 in the Details of Extrude1. Change the extrusion depth to 15 mm in the field
of FD1, Depth and click Generate. A solid body is created as shown below.
278
Finite Element Modeling and Simulation with ANSYS Workbench
Step 7: Launch the Steady-State Thermal Program
Double-click on the Model cell to launch the Steady-State Thermal program. Click on
the Solid under Geometry in the Outline tree. In the Details of “Solid,” click to the
right of the Material Assignment field and select Aluminum from the drop-down
menu.
Step 8: Generate Mesh
Right click on Mesh in the Project Outline. Select Insert and then Sizing from the context menu. In the Details of “Face Sizing,” enter “5e-4 m” for the Element Size. Click
on the top, bottom surfaces, and the side walls of the guitar in the Graphics window and apply the five faces to the Geometry selection.
Thermal Analysis
279
In the Details of “Mesh,” select Dropped for the Element Midside Nodes under the
Advanced option. This helps reduce the total number of nodes to an acceptable
level not exceeding the requested resources of educational licenses. Right-click on
Mesh and select Generate Mesh.
Step 9: Apply Boundary Conditions
Right-click on Steady-State Thermal(A5). Choose Insert and then Heat Flux from the
context menu. Apply a heat flux of 1000 W/m−2 to the base of the heat sink.
280
Finite Element Modeling and Simulation with ANSYS Workbench
Right-click on Steady-State Thermal(A5). Choose Insert and then Convection from the
context menu. In the Details of “Convection,” enter 30 W/(m2°C) for Film Coefficient
and 28°C for Ambient Temperature to all surfaces (a total of 29 faces) except for the
base of the heat sink.
Step 10: Solve and Retrieve Results
Right-click on Solution (A6) in the Outline, and insert Temperature, Total Heat Flux, and
Directional Heat Flux to the solution outline. In the Details of “Directional Heat Flux,”
set the Orientation to Y-axis. Right-click on Solution (A6) and click Solve.
Thermal Analysis
Click on Temperature in the Outline to review the temperature distribution.
Click on Total Heat Flux to display the heat flux with directional arrows.
281
282
Finite Element Modeling and Simulation with ANSYS Workbench
Click on Directional Heat Flux to review the heat flux isolines along Y-axis.
Part B: Transient Thermal Analysis
Step 1: Add a Transient Thermal Analysis System
Drag the Transient Thermal icon from the Analysis Systems Toolbox window and drop it
onto the Solution cell of the highlighted Steady-State system in the Project Schematic.
This creates a Transient Thermal system that shares data with Steady-State Thermal
system. The temperature distribution from the Steady-State Thermal analysis is
now the initial temperature for the Transient Thermal analysis. If the initial temperature is uniform for the Transient Thermal analysis, then this data sharing is
not needed.
Thermal Analysis
283
Step 2: Add Material Properties
Double-click on Engineering Data of Steady-State Thermal. Add a Density of 2800 kg/
m3 and Specific Heat of 870 J/(kg K) to the Properties of Aluminum. Click Return to
Project.
Step 3: Set Up Transient Thermal Analysis
Double-click on the Setup cell of the Transient Thermal system to launch the Multiple
Systems–Mechanical program. Click Yes on the pop-up window to read the modified upstream data.
Select Analysis Settings from the Outline tree. In the Details of “Analysis Settings”, under
Step Controls, set the Step End Time to 180s. Change the Auto Time Stepping to On
from Program Controlled. Change Defined By to Time.
284
Finite Element Modeling and Simulation with ANSYS Workbench
The default values for the initial and the maximum time steps are small for this
model. Set the Initial Time Step to 0.1. Set the Minimum Time Step to 0.05. Set the
Maximum Time Step to 5.
A small time step will help increase the accuracy of the model and also produce
enough result steps so the animation will have smooth transition between solution steps.
Step 4: Apply Cycling Heat Flux
Right-click on Transient Thermal (B5). Choose Insert and then Heat Flux from the context menu. In the Details of “Heat Flux,” change Magnitude to Tabular Data, and
apply the heat flux to the bottom surface of the heat sink as shown below.
Click on
tabular
data.
Thermal Analysis
285
Click on Tabular Data on the right edge of the Graphics window, and then click on the
push pin labeled AutoHide to display the Tabular Data window as shown below.
Enter the following values in the Tabular Data table.
Step 5: Apply Convective Boundary Condition
Right-click on Transient Thermal (B5). Choose Insert and then Convection from the context menu. In the Details of “Convection,” enter 30 W/(m2 °C) for Film Coefficient and
28°C for Ambient Temperature to all surfaces (a total of 29 faces) except for the base
of the heat sink.
286
Finite Element Modeling and Simulation with ANSYS Workbench
Step 6: Solve and Retrieve Results
Right-click on Solution (B6) in the Outline, and insert Temperature, Total Heat Flux, and
Directional Heat Flux. In the Details of “Directional Heat Flux,” set the Orientation to
Y-axis. Right-click on Solution (B6) and click Solve.
Click on Temperature in the Outline to review the distribution. To show Graph, click
on Graph on the right side of the Graphics window, and then click on the push pin
labeled AutoHide.
Click on Total Heat Flux to display the heat flux with directional arrows.
Thermal Analysis
287
Click on Directional Heat Flux to review the heat flux isolines along Y-axis.
Modeling tips: In the case study, an end time of 180 s is specified in the transient
setup. By default, the last set of results (solution at 180 s) from the transient analysis
is used for graphics window display. To display results at a different time point, for
example, temperature distribution at 40 s, you may change the Display Time from Last
to 40s in the Details of “Temperature.” Then right click on Solution (B6) and select Evaluate
All Results. A result at the specified time will be displayed at the Graphics window as
shown below.
288
Finite Element Modeling and Simulation with ANSYS Workbench
Part C: Thermal Stress Analysis
Step 1: Add a Static Structural Analysis System
Drag the Static Structural icon from the Analysis Systems Toolbox window and drop
it onto the Solution cell of the highlighted Steady-State system in the Project
Schematic.
This creates a Static Structural system that shares data with the Steady-State Thermal
system as shown below. The temperature distribution from the Steady-State
Thermal analysis is now the load input for the Static Structural analysis. If a uniform temperature is specified as a load for the Static Structural analysis, then data
sharing of the steady-state thermal solution is not needed.
Thermal Analysis
289
Step 2: Add Material Properties
Double-click on Engineering Data of Steady-State Thermal. Add Young’s Modulus
of 70 GPa, Poisson’s ratio of 0.3, and Isotropic Instantaneous Coefficient of Thermal
Expansion of 2.2E-5 1/°C to the Properties of Aluminum as shown below. Click
Return to Project.
Step 3: Set Up Static Structural Analysis
Double-click on the Setup cell of the Static Structural system to launch the Multiple
Systems–Mechanical program. Click Yes on the pop-up window to read the modified upstream data.
Note an Imported Load item is automatically added to Static Structural (C5) in the
Outline tree. Right-click on Static Structural (C5) and insert a Fixed Support to the
Outline. Apply the fixed support to the bottom face of the heat sink.
290
Finite Element Modeling and Simulation with ANSYS Workbench
Step 4: Solve and Retrieve Results
Right-click on Solution (C6) in the Outline, and insert Total Deformation and Equivalent
Stress to the outline. Then right-click on Solution (C6) and click Solve.
Click on Total Deformation in the Outline to review displacement results.
Click on Equivalent Stress in the Outline to review von Mises stress results.
Thermal Analysis
291
Modeling tips: To apply a uniform temperature load to the heat sink, you may add a
Static Structural system that shares data with the Steady-State Thermal system at the model
and above levels. And then insert a constant Thermal Condition load to the heat sink in the
Static Structural analysis.
To do this, first drag the Static Structural icon from the Analysis Systems Toolbox window
and drop it onto the Model cell of the highlighted Steady-State system in the Project Schematic.
This creates a schematic where Static Structural is sharing data with Steady-State Thermal
at the model and above levels as shown below.
292
Finite Element Modeling and Simulation with ANSYS Workbench
Next, add Young’s Modulus, Poisson’s ratio, and Coefficient of Thermal Expansion
data to the Properties of Aluminum in the Steady-State Thermal analysis template following Step 2 of Part C. Set up the Static Structural analysis by following Step 3 of Part C.
Insert a Thermal Condition load to the Static Structural (C5) in the Outline tree and apply
a Fixed Support to the base of the heat sink. The total deformation and the equivalent
stress distributions of the heat sink with a temperature increase of 10°C can be obtained
as follows.
C: Static structural
Total deformation
Type : Total deformation
Unit: m
Time: 1
C: Static structural
Equivalent stress
Type : Equivalent (von Mises) stress
Unit: Pa
Time: 1
3.2799e–6 Max
4.9202e7 Max
0 Min
12,114 Min
2.9154e–6
2.551e–6
2.1866e–6
1.8222e–6
1.4577e–6
1.0933e–6
7.2886e–7
3.6443e–7
4.3736e7
3.8271e7
3.2805e7
2.734e7
2.1874e7
1.6409e7
1.0943e7
5.4776e6
9.5 Summary
In this chapter, we discussed the governing equations for heat conduction problems and
the FEA formulation. Thermal stresses due to changes of temperatures in structures are
also discussed and the effects of constraints of the structures on the thermal stresses are
emphasized.
9.6 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1.
2.
3.
4.
Understand the FE formulations for heat conduction and thermal stress analysis
Conduct heat conduction analysis of structures using FEA
Conduct thermal stress analysis of structures using FEA
Perform heat conduction and thermal stress analysis using ANSYS Workbench
293
Thermal Analysis
PROBLEMS
9.1 Study the heat conduction problem in a simple annular region shown below,
using the FEA. Assume a = 1 m, b = 2 m, Ta = 100°C, and heat flux Qb = 200 W/m2.
Using structural steel with thermal conductivity k = 60.5 W/(m°C) for the region,
determine the temperature field and heat flux in this region and compare your
FEA results with the analytical solution.
a
b
O
V
Sa
Sb
9.2 For the glass cup shown in the figure below, determine the thermal stresses
when the inner surfaces of the cup experience a temperature change from 20°C
to 60°C while all other surfaces are kept at 20°C. For glass, use Young’s modulus
E = 70 GPa, Poisson’s ratio ν = 0.17, thermal conductivity k = 1.4 W/(m°C), and coefficient of thermal expansion α = 8.0 × 10−6/°C.
All dimensions are in millimeters.
40.000
4.000
100.000
4.000
26.000
30.000
294
Finite Element Modeling and Simulation with ANSYS Workbench
9.3 Determine the temperature and heat flux distributions inside the following heat
transfer pipe with internal ribs. A bulk temperature of 80°C and 20°C with a film
coefficient of 100 W/(m2°C) and 30 W/(m2°C) is specified for the interior and exterior convective heat transfer, respectively. Assume that the pipe is 30 cm long with
a thermal conductivity of 230 W/(m K) and that the two ends of the pipe are perfectly insulated.
All dimensions are in centimeters.
0.450
R5.500
R5.000
0.450
2.500
2.500
0.450
2.500
2.500
0.450
9.4 A 10 cm thick perforated concrete slab is shown below. The concrete has a thermal conductivity k = 1.0 W/(m K), Young’s modulus E = 29 GPa, Poisson’s ratio
ν = 0.15, and thermal expansion coefficient α = 11 × 10−6/°C. The bulk temperature of the fluid inside the circular perforated holes is 60°C with a film coefficient of 100 W/(m 2°C). Suppose the four sides of the concrete slab are kept at
a constant temperature of 20°C. The front and back faces of the perforated slab
are perfectly insulated. (1) Determine the temperature distribution in the slab.
(2) Find the thermally induced deformation and stresses in the slab if the four
sides are fixed.
All dimensions are in centimeters.
Side faces
6.000
D5.000
25.000
Front face
6.000
25.000
295
Thermal Analysis
9.5 The following thermal bimorph is made of Silicon with a thermal conductivity
k = 130 W/(m°C), Young’s modulus E = 200 GPa, Poisson’s ratio ν = 0.27, and thermal expansion coefficient α = 2.6 × 10−6/°C. It is 1 mm thick with fixed temperatures of 60°C and 10°C, respectively, at the leftmost ends of its hot arm and cold
arm, as shown below. Suppose the bimorph has a uniform internal heat generation at 100 W/m3, and ignore the effects of convective heat transfer. (1) Determine
the steady-state temperature distribution in the thermal bimorph. (2) Determine
the thermally induced deformation and stresses in the bimorph if the two arms
are fixed on their left ends.
Hot arm
Thot = 60°C
Cold arm
Tcold = 10°C
6.000
All dimensions are in millimeters.
3.000
2.000
2.000
5.000
12.000
2.000
3.000
9.6 The heat sink shown below has a square base of size 22 mm × 22 mm × 2 mm
and consists of an array of square pin-fins, each of size 2 mm × 2 mm × 8 mm.
The pin-fins are separated by a gap distance of 4 mm in each direction. The heat
sink is initially at a temperature of 28°C. Suppose a heat flux input is prescribed
on the bottom face of the heat sink. The heat flux is of a symmetrical square wave
pattern with a period of 60 s (equal time for each half-cycle) and peak magnitudes transitioning from 800 W/m 2 to 0 W/m 2. A fan forces air over all surfaces
of the heat sink except for the bottom face, where the heat flux is prescribed. The
surrounding air is 28°C with a film coefficient of 30 W/(m 2°C). Assume that the
heat sink is made of aluminum with thermal conductivity k = 170 W/(m K), density ρ = 2800 kg/m3, specific heat c = 870 J/(kg K), Young’s modulus E = 70 GPa,
Poisson’s ratio ν = 0.3, and thermal expansion coefficient α = 22 × 10−6/°C. Study
the transient thermal response of the heat sink during the first three cycles.
296
Finite Element Modeling and Simulation with ANSYS Workbench
9.7 The electric kettle base shown below has a thickness of 0.5 cm and is
made of structural steel with thermal conductivity k = 60.5 W/(m K), density ρ = 7850 kg/m3, and specific heat c = 434 J/(kg K). Suppose a heat flux condition is defined through a serpentine heating element embedded at the bottom
of the kettle base. Except for the imprinted heating element face, all other faces
of the kettle base are subjected to convective heat transfer with a bulk air temperature of 40°C and a film coefficient of 80 W/(m2°C). (1) Study the steady-state
thermal response of the kettle base if a constant heat flux of 800 W/(m2°C) is
applied through the heating element. (2) Suppose the heating element is cycled
on and off with a period of 60 s (30 s on-time followed by 30 s off-time). During
on-time, the heat flux is of magnitude 1000 W/(m2°C). Study the transient thermal response of the kettle base for a period of 120 s using the steady-state solution as the initial condition.
Isometric view of the two sides of the kettle base
Heating element is shown as
an imprinted face on the
bottom of the kettle base.
0.500
1.000
R6.000
R1.500
2.000
1.500
R10.000
All dimensions are in
centimeters.
8.000
10.000
R1.250
2.000
297
Thermal Analysis
9.8 The air to water heat exchanger shown below is made of copper with thermal
conductivity k = 400 W/(m K), Young’s modulus E = 100 GPa, Poisson’s ratio
ν = 0.3, and thermal expansion coefficient α = 18 × 10−6/°C. The exterior surfaces
are in contact with cold water with a film coefficient of 30 W/(m2°C) and a bulk
temperature of 20°C. The interior surfaces are in contact with hot air with a film
coefficient of 100 W/(m2°C) and a bulk temperature of 80°C. (1) Determine the
steady-state thermal response of the heat exchanger. (2) Suppose the two annulus
faces at the ends of the heat exchanger are fixed. Determine the thermal deformation and stresses induced in the exchanger.
All dimensions are in millimeters.
40.000
R5.000
20.000
R3.000
R10.000
10.000
R10.000
20.000
9.9 Thermal expansion mismatch is of great concern in high-density semiconductor
devices. The model shown below consists of a copper dye on a silicon substrate.
Suppose the dye and substrate are perfectly bounded and the substrate is fixed
on the bottom face. Determine the thermally induced deformation and stresses if
the dye is assigned a uniform temperature of 60°C and the substrate is assigned
a uniform temperature of 40°C. For copper, use Young’s modulus E = 100 GPa,
Poisson’s ratio ν = 0.3, and thermal expansion coefficient α = 18 × 10−6/°C, and for
silicon, use E = 200 GPa, ν = 0.27, and α = 2.6 × 10−6/°C.
All dimensions are in millimeters.
Die thickness = 2 mm
Substrate thickness = 2 mm
40.000
20.000
Copper die
Silicon substrate
20.000
40.000
298
Finite Element Modeling and Simulation with ANSYS Workbench
9.10 The following square platform with two arms has a thickness of 1 mm and is
made of silicon with a thermal conductivity k = 130 W/(m°C), Young’s modulus E = 200 GPa, Poisson’s ratio ν = 0.27, and thermal expansion coefficient
α = 2.6 × 10−6/°C. Ignore the effects of convective heat transfer. (1) Determine the
temperature distribution in the platform when the temperatures of the upper
and lower ends are set uniform at 40°C and 20°C, respectively, as shown below.
(2) Suppose the ends of the two arms are fixed. Determine the thermally induced
deformation and stresses in the platform.
Tupper = 40°C
All dimensions are in centimeters.
1.000
12.000
1.000
1.000
10.000
1.000
Tlower = 20°C
1.000
10.000
12.000
1.000
10
Introduction to Fluid Analysis
10.1 Introduction
Flow simulations are widely used in engineering applications ranging from flow around
airplane wings and hydraulic turbines to flow in blood vessels and other circulatory systems (see Figure 10.1). We may gain a better understanding of the motion of fluid around
objects as well as the fluid behavior in complex circulatory systems by conducting fluid
analysis. Computational fluid dynamics (CFD) simulation complements experimental testing, helps reduce cost and turnaround time for design iterations, and has become an indispensable tool whenever practical design involving fluids is required. In this chapter, we
will discuss fluid analysis using ANSYS® Workbench.
10.2 Review of Basic Equations
We begin by briefly reviewing the fundamentals of fluid mechanics.
10.2.1 Describing Fluid Motion
In fluid dynamics, the motion of a fluid is mathematically described using physical quantities such as the flow velocity u, flow pressure p, fluid density ρ, and fluid viscosity ν. The
flow velocity or flow pressure is different at a different point in a fluid volume. The objective of fluid simulation is to track the fluid velocity and pressure variations at different
points in the fluid domain.
10.2.2 Types of Fluid Flow
There are many different types of fluid flow. A flow can be compressible (ρ ≠ constant)
or incompressible (ρ = constant), viscous (ν ≠ 0) or inviscid (ν = 0), steady (∂u/∂t = 0) or
unsteady (∂u/∂t ≠ 0) and laminar (streamline) or turbulent (chaotic). Furthermore, a fluid
can be Newtonian (if the viscosity depends only on temperature and pressure, not on
forces acting upon it; in other words, shear stress is a linear function of the fluid strain rate)
or non-Newtonian (if the viscosity depends on forces acting upon it, i.e., shear stress is a
nonlinear function of the fluid strain rate).
299
300
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 10.1
Examples of CFD: (a) airflow around a tractor (Courtesy ANSYS, Inc., http://www.ansys.com/Industries/
Automotive/Application+Highlights/Body), and (b) streamlines inside a combustion chamber (Courtesy
ANSYS, Inc., http://www.edr.no/en/courses/ansys_cfd_advanced_modeling_ reacting_flows_and_combustion_in_ans ys_fluent).
10.2.3 Navier–Stokes Equations
For the purpose of this chapter, we limit ourselves to the study of incompressible
Newtonian flows. All fluids are compressible to some extent, but we may consider most
common liquids as incompressible, whose motion is governed by the following Navier–
Stokes (N–S) equations:
∇p
∂u
= −u∇u + ν∇ 2u −
+f
∂t
ρ
(10.1)
which shows that the acceleration ∂u/∂t of a fluid particle can be determined by the combined effects of advection u∇u, diffusion ν∇ 2u, pressure gradient ∇p/ρ, and body force f.
The N–S equations can be derived directly from the conservation of mass, momentum,
and energy principles. Note that for each particle of a fluid field we have a set of N–S
equations. A particle’s change in velocity is influenced by how the surrounding particles
are pushing it around, how the surrounding resists its motion, how the pressure gradient
changes, and how the external forces such as gravity act on it [15].
In 3-D Cartesian coordinates, the N–S equations become:
∂u
∂ 2 u ∂ 2 u ∂ 2 u ∂p
∂u
∂u
∂u
ρ
+u
+v
+ w = ν 2 + 2 + 2 −
+ fx
∂x
∂y
∂z
∂y
∂z ∂x
∂t
∂x
(10.2)
∂v
∂ 2 v ∂ 2 v ∂ 2 v ∂p
∂v
∂v
∂v
ρ
+u
+v
+ w = ν 2 + 2 + 2 −
+ fy
∂x
∂y
∂z
∂y
∂z ∂y
∂t
∂x
(10.3)
∂w
∂ 2 w ∂ 2 w ∂ 2 w ∂p
∂w
∂w
∂w
ρ
+u
+v
+w
=
ν
∂x 2 + ∂y 2 + ∂z 2 − ∂z + f z
∂x
∂y
∂z
∂t
(10.4)
where u, v, and w are components of the particle’s velocity vector u.
In CFD modeling, the N–S equations for particle motion are numerically solved, along
with specified boundary conditions, on a 3-D grid that represents the fluid domain to be
301
Introduction to Fluid Analysis
analyzed. For more details on numerical solution techniques adopted in CFD, please refer
to the theory guide in ANSYS online documentation.
10.3 Modeling of Fluid Flow
Practical aspects of CFD modeling are discussed next. The topics include fluid domain
specification, meshing, boundary condition assignments, and solution visualization.
10.3.1 Fluid Domain
A fluid domain is a continuous region with respect to the fluid’s velocity, pressure, density,
viscosity, and so on. Figure 10.2 illustrates examples of an internal flow and an external
flow. For an internal flow (see Figure 10.2a), the fluid domain is confined by the wetted
surfaces of the structure in contact with the fluid. For an external flow (see Figure 10.2b),
the fluid domain is the external fluid region around the immersed structure.
10.3.2 Meshing
In CFD analysis, mesh quality has a significant impact on the solution time and accuracy
as well as the rate of convergence. A good mesh needs to be fine enough to capture all relevant flow features, such as the boundary layer and shear layer and so on (see Figure 10.3),
without overwhelming the computing resources. A good mesh should also have smooth
and gradual transitions between areas of different mesh density, to avoid adverse effect on
convergence and accuracy.
10.3.3 Boundary Conditions
Appropriate boundary conditions are required to fully define the flow simulation, as the
flow equations are solved subject to boundary conditions. The common fluid boundary
conditions include the inlet, outlet, opening, wall, and symmetry plane [16]. An inlet condition is used for boundaries where the flow enters the domain. An outlet condition is for
where the flow leaves the domain. An opening condition is used for boundaries where the
fluid can enter or leave the domain freely. A wall condition represents a solid boundary of
(a)
Internal flow
(b)
External flow
FIGURE 10.2
Flow region definition: (a) internal flow and (b) external flow.
302
Finite Element Modeling and Simulation with ANSYS Workbench
Flow
Inadequate mesh
Better mesh
FIGURE 10.3
The resolution of a mesh needs to adequately represent the flow feature.
the flow model. For fixed walls, this typically means a no-slip boundary condition for the
flow velocity. A symmetry plane condition is used for planes exhibiting both geometric
and flow symmetry.
A pressure boundary is a boundary where the flow pressure is defined. A velocity
boundary is a boundary where the velocity profile is given. Figure 10.4 gives two simulated airflow streamline plots for a city block. The boundary conditions for (a) include a
velocity inlet on the left edge, a pressure outlet on the right edge, and wall conditions on all
buildings (depicted in gray lines). The boundary conditions for (b) include pressure inlets
on the four sides of the city block, a pressure outlet at the city’s circular center, and wall
conditions on all buildings. As we can see from Figure 10.4, different boundary conditions
result in different flow patterns. Assigning realistic boundary conditions that truly reflect
the flow conditions is crucial in CFD analysis.
10.3.4 Solution Visualization
Once a problem is setup with the fluid domain meshed and the boundary conditions specified, it can be submitted to the solver for computation of a solution. Flow motion can be
described by plotting the flow variables on a section plane or in 3-D space. The commonly
used plots in CFD are contour plots of pressure, shear strain rate and turbulence kinetic
(a)
(b)
0
5.00e+004
1.00e+005 (m)
0
5.00e+004
1.00e+005 (m)
FIGURE 10.4
Airflow over a city block: (a) crosswind blowing from left to right; (b) a circular low-pressure area at the city center.
303
Introduction to Fluid Analysis
(a)
(b)
Shear strain rate
Contour 1
7.329e-002
6.600e-002
5.871e-002
5.142e-002
4.412e-002
3.683e-002
2.954e-002
2.225e-002
1.495e-002
7.661e-003
3.685e-004
[sˆ–1]
Pressure
Contour 1
3.820e-005
3.105e-005
2.390e-005
1.674e-005
9.590e-006
2.438e-006
–4.715e-006
–1.187e-005
–1.902e-005
–2.617e-005
–3.333e-005
[Pa]
(d)
(c)
Turbulence kinetic energy
Contour 2
6.714e-006
6.052e-006
5.390e-006
4.728e-006
4.066e-006
3.404e-006
2.742e-006
2.080e-006
1.418e-006
7.558e-007
9.375e-008
[mˆ2 sˆ–2]
Velocity
Streamline 1
7.669e-003
5.763e-003
3.858e-003
1.952e-003
4.635e-005
[m sˆ–1]
(e)
Velocity
Vector 1
7.669e-003
5.752e-003
3.835e-003
1.917e-003
0.000e-000
[m sˆ–1]
FIGURE 10.5
Solution plots of fluid flow past a single cylinder (a half-model using symmetry). (a) Pressure distribution; (b)
shear strain rate distribution; (c) turbulence kinetic energy distribution; (d) streamline; and (e) velocity vector
plot.
energy distributions, streamline plot, and vector plot of velocity field. Figure 10.5 shows
the related solution plots of a flow passing over a single cylinder. A symmetric half-model
is used in the example, with the flow pressure, velocity, shear stress, turbulence intensity,
and particle trajectories plotted as shown above.
10.4 Case Studies with ANSYS Workbench
Problem Description: The aerodynamic performance of vehicles can be improved by
utilizing computational fluid dynamics simulation. In this case study, we conduct fluid
304
Finite Element Modeling and Simulation with ANSYS Workbench
analysis of the air flow passing through a truck. Assume air at room temperature of 25°C
for the flow field with an air velocity of 40 km/h blowing from left to right. Use nonslip
boundary conditions along the walls of the truck and the ground surface. Find the airflow
pattern as well as the pressure and velocity distributions of the flow field around the truck.
R0.200
3.000
1.200
2.000
1.200
R0.500
R0.400
0.800
R0.400
2.400
All dimensions are in meters.
Flow field: Incompressible air at 25°C.
Boundary conditions: Air is blowing from left to right at a velocity of 40 km/h.
Nonslip boundaries along the walls of the truck and the ground surface.
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “CFD_Truck.wbpj.”
Step 2: Create a Fluid Flow (CFX) Analysis System
Drag the Fluid Flow (CFX) icon from the Analysis Systems Toolbox window and drop it
inside the highlighted green rectangle in the Project Schematic window.
Introduction to Fluid Analysis
305
Step 3: Launch the DesignModeler Program
Double-click the Geometry cell to launch DesignModeler, and select “Meter” in the
Units pop-up window.
Step 4: Create a Truck Model
Create a Sketch—Click on Sketching. Select Draw and draw a 2D profile as shown
below.
306
Finite Element Modeling and Simulation with ANSYS Workbench
Create an Extruded Body—Switch to the Modeling tab and click on Extrude. In the
Details of Extrude1, use Sketch1 as the Base Object and change the Direction to BothSymmetric. Set the extrusion depth to 0.8 m in the field of FD1, Depth, and click
Generate to create a solid body.
Create Sketch of Two Wheels—Click on the New Plane button to create a new plane
(Plane4) on the front face of the truck. Then click on the New Sketch button to create a new sketch (Sketch2) on the plane as shown below.
Then click on new sketch button.
Click on new
plane button
Introduction to Fluid Analysis
307
Create Two Solid Wheels—Switch to the Modeling tab and click on Extrude. In the
Details of Extrude2, use Sketch2 as the Base Object and change the Direction to BothSymmetric. Set FD1, Depth to 0.1 m. Click Generate to create two wheels.
Create the Other Two Wheels—Repeat the previous two steps to create two wheels
on the other side of the truck model as shown below.
308
Finite Element Modeling and Simulation with ANSYS Workbench
Step 5: Create a Box Enclosure
Select Enclosure from the drop-down menu of Tools to create a box enclosure.
In the Details of Enclosure1, enter the cushion dimensions as shown below, and click
Generate.
Introduction to Fluid Analysis
309
In the Tree Outline, right click on the first Solid body, and select Suppress Body. The
first solid body represents the truck. When suppressed, the truck body will not be
meshed later on.
From the Tree Outline click on the second Solid body, which represents the enclosing
fluid domain. Ensure that Fluid is selected for the Fluid/Solid field in the Details of
Body.
310
Finite Element Modeling and Simulation with ANSYS Workbench
Step 6: Generate Mesh
Double-click on the Mesh cell to launch the Meshing program.
Click on Mesh in the Outline. In the Details of “Mesh,” set the Relevance Center to Medium
under Sizing, and the Element Midside Nodes to Dropped. This helps reduce the total
number of nodes to an acceptable level not exceeding the requested resources of
educational licenses. Click on Update to generate the mesh.
Introduction to Fluid Analysis
The following figure shows the generated mesh for the fluid domain.
Step 7: Set Up Flow Domain
Double-click on the Setup cell to launch the Fluid Flow (CFX)—CFX-Pre program.
311
312
Finite Element Modeling and Simulation with ANSYS Workbench
Double-click on Analysis Type in the Outline tree.
In the Details of Analysis Type in Flow Analysis 1, set the Analysis Type to Steady State.
Click Ok to exit the menu.
Introduction to Fluid Analysis
313
Next, double-click on Default Domain in the Outline tree.
In the Details of Default Domain in Flow Analysis 1, set the Domain Type to Fluid Domain,
Material to Air at 25°C, and Reference Pressure to 1[atm]. Click Ok to exit the menu.
314
Finite Element Modeling and Simulation with ANSYS Workbench
Step 8: Set Up Boundary Conditions
Add Inlet—Right-click on Default Domain in the Outline. Select Insert and then Boundary.
Change the Name to Inlet in the pop-up menu shown below. Click Ok.
In the Details of Inlet in Default Domain in Flow Analysis 1, select the face highlighted
below for the Location of the inlet under the tab of Basic Settings.
Introduction to Fluid Analysis
315
Click on the Boundary Details tab, and set the Normal Speed to 40 [Km Hr ^ −1]. Click OK.
Add Outlet—Right-click on Default Domain in the Outline. Select Insert and then
Boundary. In the pop-up menu shown below, set the Name to Outlet. Click Ok.
In the Details of Outlet in Default Domain in Flow Analysis 1, select the face highlighted
below for the Location of the outlet under the tab of Basic Settings.
316
Finite Element Modeling and Simulation with ANSYS Workbench
Click on the Boundary Details tab, and set the Relative Pressure to 0 [Pa]. Click Ok.
Add Wall—Right-click on Default Domain in the Outline. Select Insert and then
Boundary. In the pop-up menu shown below, set the Name to Wall. Click Ok.
In the Details of Wall in Default Domain in Flow Analysis 1, click on the button to the
right of the Location field under the tab of Basic Settings.
Click on the button to
select faces
Introduction to Fluid Analysis
317
In the pop-up menu of Selection Dialog, select multiple faces (all surfaces of the truck
and the ground surface) with Shift-Click. Click Ok to accept the selection.
Switch to the tab of Boundary Details, select No Slip Wall as the Option for Mass and
Momentum, and Smooth Wall for Wall Roughness. Click Ok to exit the menu.
318
Finite Element Modeling and Simulation with ANSYS Workbench
Add Opening—Right-click on Default Domain in the Outline. Select Insert and then
Boundary. In the pop-up menu shown below, set the Name to Opening. Click Ok.
In the Details of Opening in Default Domain in Flow Analysis 1, click on the button to the
right of the Location field under the tab of Basic Settings.
Click on the button to
select faces
In the pop-up menu of Selection Dialog, select the three side faces with Ctrl-Click as
shown below. Click Ok to accept the selection.
Introduction to Fluid Analysis
319
Switch to the tab of Boundary Details, set the Relative Pressure to 0 [Pa]. Click Ok to exit.
320
Finite Element Modeling and Simulation with ANSYS Workbench
After completion, the finished domain boundaries should look like the following.
Exit CFX-Pre.
Step 9: Start a Solution
Double-click on the Solution cell to obtain a solution of the fluid flow problem.
Introduction to Fluid Analysis
321
In the pop-up menu of Define Run, click the Start Run button to launch the CFX-Solver.
When a menu pops up to indicate that the Solver Run has finished normally, click Ok.
Then exit the CFX-Solver Manager program.
322
Finite Element Modeling and Simulation with ANSYS Workbench
Step 10: Retrieve Results
Double-click on the Results cell to visualize the results.
Plot Velocity Vector—Right-click User Locations and Plots in the Outline, and select
Insert and then Vector. Accept the default name Vector 1 in the pop-up menu. In the
Details of Vector 1, select All Domains for Domains, Default Domain for Locations and
Velocity for Variable, and click Apply. Check Default Legend View 1 in the Outline to
display the legend.
Plot Wall Pressure Contour—Right-click User Locations and Plots in the Outline, and
select Insert and then Contour. Accept the default name Contour 1 in the pop-up
Introduction to Fluid Analysis
323
menu. In the Details of Contour 1, select All Domains for Domains, Wall for Locations
and Pressure for Variable, and click Apply. Check Default Legend View 1 and Contour1
in the Outline to view the pressure distribution on the wall.
Plot Streamline—Right-click User Locations and Plots in the Outline, and select Insert
and then Streamline. Accept the default name Streamline 1 in the pop-up menu.
In the Details of Streamline 1, select 3D Streamline for Definition, All Domains for
Domains, Inlet for Start From, and Velocity for Variable, and click Apply. Check Default
Legend View 1, Streamline 1, and Wireframe in the Outline to view streamlines.
324
Finite Element Modeling and Simulation with ANSYS Workbench
Modeling tips: Follow the steps below to visualize flow results on a section plane. First,
right-click User Locations and Plots in the Outline, then select Insert and Location and then
Plane in the context menu. Accept the default name Plane 1 in the pop-up menu. In the
Details of Plane 1, select All Domains for Domains, XY Plane for Method, and 0.0 [m] for Z, and
click Apply. Check Plane 1 and Wireframe in the Outline to view the created section plane.
To generate a contour plot on the plane, right-click User Locations and Plots, and select
Insert and then Contour. Accept the default name Contour 2 in the pop-up menu. In the
Details of Contour 2, select All Domains for Domains, Plane 1 for Locations, and Velocity for
Variable, and click Apply. Check Default Legend View 1, Contour2, and Wireframe in the Outline
to retrieve the velocity distribution on the section plane.
Introduction to Fluid Analysis
325
To generate a streamline plot on the plane, right-click User Locations and Plots, and select
Insert and then Streamline. Accept the default name Streamline 2 in the pop-up menu. In the
Details of Streamline 2, select All Domains for Domains, Plane 1 for Start From, Rectangular Grid
for Sampling, 0.03 for Spacing, and click Apply. Check Default Legend View 1, Streamline2, and
Wireframe in the Outline to view the streamlines on the section plane.
10.5 Summary
In this chapter, we briefly discussed the governing equations in fluid dynamics, main
variables of concern, boundary conditions, and procedures in conducting CFD modeling
and analysis. A case study using ANSYS Workbench is demonstrated using a truck model.
10.6 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1. Understand the unique nature of fluid dynamics problems, including its governing equations, variables, and boundary conditions
2. Know how to create quality mesh for CFD analysis
3. Know how to apply the boundary conditions for CFD analysis
4. Perform CFD modeling and analysis of flows in complex domains using ANSYS
Workbench
326
Finite Element Modeling and Simulation with ANSYS Workbench
PROBLEMS
10.1 Conduct fluid analysis of the water flow passing through a U-tube. The tube has
an inner and outer diameter of 4 and 4.2 cm, respectively. Assume nonslip boundary conditions on the surface walls. The inlet flow velocity is 5 mm/s and the
outlet flow velocity is 4 mm/s. Determine the pressure and velocity distributions
of the interior flow field.
R5.000
All dimensions are in centimeters.
Inlet
10.000
10.000
Outlet
10.2 Determine the pressure and velocity distributions of the flow of air over a solid
half-cylinder. The 20 cm long half-cylinder has a radius of 5 cm, and is placed on
the ground. The upstream air has a velocity of 5 mm/s, and the air in the downstream passageway is at atmospheric pressure. Assume nonslip boundary conditions at the air–solid interface and the air is at a constant room temperature of
25°C.
Flow direction
R5.000
All dimensions are in
centimeters.
327
Introduction to Fluid Analysis
10.3 Determine the pressure and velocity distributions of the flow of air over a solid
sphere. The sphere has a radius of 5 cm. The upstream air has a velocity of 5 mm/s,
and the air in the downstream passageway is at atmospheric pressure. Assume
nonslip boundary conditions at the spherical surface and the air is at a constant
room temperature of 25°C.
R5.000
All dimensions are in
centimeters.
Flow direction
10.4 The heat sink shown below has a square base of size 22 mm × 22 mm × 2 mm and
consists of an array of square pin-fins, each of size 2 mm × 2 mm × 8 mm. The
pin-fins are separated by a gap distance of 4 mm in each direction. Determine the
pressure and velocity distributions of the flow of air over the heat sink at room
temperature of 25°C. Suppose the square base of the heat sink is in contact with
the ground. The upstream air has a velocity of 10 mm/s, and the air in the downstream passageway is at atmospheric pressure.
Flow direction
328
Finite Element Modeling and Simulation with ANSYS Workbench
10.5 Conduct fluid analysis of the air flow passing through an S-pipe. Assume nonslip
boundary conditions along the pipe walls and the air is at a room temperature of
25°C. The velocity of the air is 5 mm/s at the inlet and the outlet is at atmospheric
pressure. Determine the pressure and velocity distributions of the interior flow field.
All dimensions are in millimeters.
R5.000
40.000
R3.000
20.000
Outlet 10.000
R10.000
R10.000
Inlet
20.000
10.6 Conduct fluid analysis of the water flow passing through an orifice. Assume
nonslip boundary conditions on the walls. The water has a uniform velocity of
15 mm/s at the inlet and the outlet is at atmospheric pressure. Determine the
pressure and velocity distributions of the flow field.
All dimensions are in centimeters.
10.000
6.000
10.000
Inlet
1.000
Outlet
10.7 Conduct fluid analysis of the water flow passing through a diffuser pipe. Assume
nonslip boundary conditions along the walls. The water has a uniform velocity
of 10 mm/s at the inlet and the outlet is at atmospheric pressure. Determine the
pressure and velocity distributions of the flow field.
All dimensions are in centimeters.
6.000
3.000
Inlet
10.000
6.000
Outlet
329
Introduction to Fluid Analysis
10.8 Conduct fluid analysis of the water flow passing through a transition pipe. Assume
nonslip boundary conditions along the walls. The water has a uniform velocity
of 25 mm/s at the inlet and the outlet is at atmospheric pressure. Determine the
pressure and velocity distributions of the flow field.
All dimensions are in centimeters.
8.000
8.000
8.000
3.000
5.000
Outlet
Inlet
10.9 Determine the pressure and velocity distributions of the flow of water inside the
following circular chamber. The chamber has a depth of 2 m. Assume nonslip
boundary conditions at the walls. The velocity of water at the inlets is 10 cm/s,
and the outlets are at atmospheric pressure.
Outlet
2.000
All dimensions are in meters.
2.000
R4.000
2.000
2.000
R1.500
Inlet
Inlet
2.000
2.000
2.000
Outlet
2.000
330
Finite Element Modeling and Simulation with ANSYS Workbench
10.10 Determine the pressure and velocity distributions of the flow of water inside the
following circular chamber. The chamber has a depth of 5 mm. Assume nonslip
boundary conditions at the walls. The velocity of water at the inlets is 5 mm/s,
and the outlets are at atmospheric pressure.
Inlet
4.000
R25.000
4.000
4.000
45.00°
R12.000
R15.000
45.00°
45.00°
Inlet
D10.000
4.000
45.00°
4.000
4.000
Outlet
All dimensions are in millimeters.
4.000
Outlet
4.000
11
Design Optimization
11.1 Introduction
Optimization is an integral part of engineering design. Traditionally, optimization tasks
have been carried out mostly by trial and error, when unexpected failure of a design to
meet certain criteria is identified. Often, a best design is not obtained after many design
iterations. Only a feasible design meeting all the requirements is created. The conventional way of changing a design when it is found to be deficient in certain criteria can be
incredibly costly and time-consuming. For today’s engineers, a more productive and costeffective practice is to use numerical optimization techniques to guide in the evaluation
of design trade-offs. Often, a best design is put forward after running simulation-based
optimization a few times. In this chapter, we will cover materials relating to the topics of
design optimization via simulation. The concepts of topology optimization, parametric
optimization, and design space exploration will be introduced, along with optimization
examples using ANSYS® Workbench.
11.2 Topology Optimization
A design’s performance should be optimized from different perspectives as the design
process evolves. In the early stage of a design, topology optimization can be used to help
designers arrive at a good initial design concept. The goal of topology optimization is to
find the ideal distribution of material within a predefined design space for a given set of
loading and boundary conditions. The regions that contribute the least to the load bearing are identified and taken out from the design to minimize the weight. As a result, an
optimal material layout is determined, from which a good design concept can be derived.
Figure 11.1 illustrates topology optimization of a bridge structure. The 3-D design space
of the bridge is shown as a solid rectangle box in Figure 11.1a. The bridge is fixed on the
bottom two edges and applied a surface load on the top face. In Figure 11.1b, an arch is
clearly suggested as the ideal layout by the topology optimization study aiming at 80%
weight reduction. Materials are removed from the least stressed regions in the simulation
model, that is, regions contributing the least to the overall stiffnesss of the structure. This
optimization study perhaps helps elucidate why the arch continues to play an important
part in bridge design after thousands of years of architectural use.
331
332
Finite Element Modeling and Simulation with ANSYS Workbench
FIGURE 11.1
Topology optimization of a bridge structure: (a) The original design space and (b) the optimized layout with an
80% weight reduction target.
11.3 Parametric Optimization
In the final design stage, a design’s performance is greatly influenced by its shape and size.
Parametric optimization can be used to help designers determine the optimal shape and
dimensions of a structure. In parametric optimization, the independent variables whose
values can be changed to improve a design are called design variables. Design variables are
usually geometric parameters such as length, thickness, or control point coordinates that
control a design’s shape. Responses of the design to applied loads are known as state variables, which are functions of the design variables. Examples of state variables are stresses,
deformations, temperatures, frequencies, and so on. The restrictions placed on the design
are design constraints. In general, parametric optimization involves minimizing an objective function of the design variables subject to a given set of design constraints [17].
For example, consider the parametric optimization of a stiffened aluminum panel with
clamped edges. Stiffened panels are suited for weight-sensitive designs and are widely
used in ship decks, air vehicles, and offshore structures. The main drawback is that they
are light-weight structures with low natural frequencies, leading to a greater risk of resonance. For the design shown in Figure 11.2, the variables that are allowed to change, that
is, the design variables, are the stiffener height h, the plate thickness t, the longitudinal
stiffener thickness tlong, and the lateral stiffener thickness tlat. To reduce the panel’s vulnerability to vibration-induced movement, its fundamental frequency f base, that is, the state
variable in the study, is to be set above, say, 20 Hz. Suppose your objective is to minimize
the panel’s overall weight. You may set up a parametric optimization study solving for the
optimum values of the four design variables so as to minimize the panel’s weight while
satisfying the design constraint of f base > 20 Hz.
11.4 Design Space Exploration for Parametric Optimization
A “black box” model shown in Figure 11.3 can be used to illustrate the parametric optimization schematic, as a direct relationship between the design variables and responses is
generally unknown. The schematic includes the use of a parametric finite element model
for finding the effects of design variables (inputs) to the responses (outputs). After multiple
response datasets are collected from finite element simulations, a mapping relationship can
333
Design Optimization
Plate thickness t
Longitudinal stiffener
thickness tlong
Lateral stiffener thickness
tlat
Stiffener height h
.
.
.
Parametric finite
element model
.
.
.
Outputs
(responses)
Inputs
(design variables)
FIGURE 11.2
A stiffened panel with clamped edges.
Optimizer: to find
optimum design
FIGURE 11.3
“Black Box” optimization schematic.
be established between the design variables and responses. The optimizer then explores
the mapping of the parametric design space to find the optimal set of design variables that
fulfill the given response criteria.
In the design space exploration, design of experiments are usually used in combination
with response surface modeling to efficiently map out a parametric design space through
simulating a minimum number of design scenarios.
11.4.1 Design of Experiments
Design of experiments (DOE) is a technique originally developed for model fitting with
experimental data. In design optimization, DOE can be used to fit the simulated response
data to mathematical equations. These equations, also referred to as response surface equations, serve as models (see Figure 11.4) to predict the responses for any combination of
design variable values.
In DOE, each design variable is viewed as one dimension in a design space. In general,
we have an n-dimensional space if there are n design variables. Each design variable has
many possible discrete values or levels. An array can be constructed by considering all the
combinations of design variable levels. In most cases, we cannot afford the response data
334
Finite Element Modeling and Simulation with ANSYS Workbench
45
40
Response
35
30
25
20
15
10
9
8
Design variable 2
7
6
1.2
5 1
1.4
1.8
1.6
2
Design variable 1
FIGURE 11.4
A 3-D response surface representation.
collection for such an array even for a fairly small value of n. As Table 11.1 shows, 16 experimental runs are needed already for four input variables (x1, x2, x3, and x4) at two levels (0
and 1). The number of simulation runs will be overwhelmingly large for multiple inputs
at many different levels.
A grid of data points of minimum density over the multivariable design space is often
used instead. Especially, central composite design (CCD) is a popular design, due to its
efficiency in providing much information in a minimum number of required numerical
TABLE 11.1
Design of Experiments: 16 Experimental Runs
for 4 Inputs at 2 Levels
Inputs
Experimental runs
x1
x2
x3
x4
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
0
0
0
0
1
1
1
1
0
0
0
0
1
1
1
1
0
0
1
1
0
0
1
1
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
335
Design Optimization
x3
x2
x1
FIGURE 11.5
A face-centered central composite design for three design variables.
experiments [17]. CCD enables an efficient construction of a second-order fitting model.
As illustrated in Figure 11.5, a typical face-centered CCD for three design variables (x1, x2,
and x3) at 3 ­levels suggests the use of 15 design points, compared to 33 possible combinations of the full design. The 15 design points are marked as back dots in Figure 11.5. Each
design point corresponds to a design scenario. By using the CCD, the maximum amount of
information can be extracted while requiring a significantly reduced number of numerical
experiments.
11.4.2 Response Surface Optimization
After the design space is sampled through an experimental design such as CCD, a response
dataset (e.g., the maximum deformation and stress results) can be readily obtained for a
design scenario through the finite element simulation. The response datasets, with each
dataset corresponding to a simulation scenario, are then used to fit the response surface
models. These models are interpolation models that can provide continuous variation of
the responses with respect to the design variables (see Figure 11.4).
In the optimizer, a designer sets up a design objective and constraints. For optimization
with multiple objectives, the relative importance of different objectives and constraints can
be specified. Using the fitted response surface models, the feasible region, that is, the region
satisfying all design constraints can be identified in the design space. The best design candidate is then determined by searching for the best available value of the objective function over the entire design space’s feasible region.
In the next section, we will use ANSYS Workbench to optimize an L-shaped structure
using the above-mentioned techniques.
11.5 Case Studies with ANSYS Workbench
Problem Description: Determine if weight reduction pockets can be generated in the
L-shaped structure shown below. The structure is 2 mm thick and is made of structural
steel. The boundary and loading conditions are specified as follows: A downward force of
300 N is applied at the bottom edge of the free end, and the top face of the L-shape is fixed.
The allowed maximum deformation in the structure is 0.3 mm. (a) Perform topology optimization to achieve 60% weight reduction. (b) Redesign the structure based on the results
from topology optimization, and conduct parametric optimization to minimize weight
subject to the deformation constraint.
336
Finite Element Modeling and Simulation with ANSYS Workbench
This top face is
fixed.
The free end
15.000
35.000
15.000
35.000
300 N
All dimensions are in mm.
Solution steps for portion (A):
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Lshape.wbpj.”
Step 2: Create a Shape Optimization Analysis System
Drag the Shape Optimization icon from the Analysis Systems Toolbox window and drop
it inside the highlighted green rectangle in the Project Schematic window.
Note: If Shape Optimization does not appear in the Toolbox of Analysis Systems, you may
have to turn on the beta option. Go to Tools > Options… > Appearance, and check the
box next to Beta Options. If beta options are on and the analysis system still is not coming up, go to the bottom of the Toolbox in Workbench and click “View All/Customize”
and make sure that there is a check mark next to “Shape Optimization (Beta).”
Step 3: Launch the DesignModeler Program
Double-click the Geometry cell to launch DesignModeler, and select “Millimeter” in the
Units pop-up window.
Design Optimization
337
Step 4: Create the Geometry
Click on the Sketching tab. Draw a sketch of the L-shape on the XY Plane, as shown
below. An entity named Sketch1 will be shown underneath XY Plane of the model’s
Tree Outline.
Extrude Sketch1 to create a 2 mm thick solid body, as shown below.
338
Finite Element Modeling and Simulation with ANSYS Workbench
Step 5: Launch the Shape Optimization Program
Double-click on the Model cell to launch the Shape Optimization program. Change the
Units to Metric (mm, kg, N, s, mV, mA).
Step 6: Generate Mesh
Click on Mesh in the Outline tree. In the Details of “Mesh,” enter “1.0 mm” for the
Element Size. Right-click on Mesh and select Generate Mesh.
Design Optimization
339
Step 7: Apply Boundary Conditions
Right-click on Shape Optimization (A5). Choose Insert and then Fixed Support from the
context menu. Click on the top face, and apply it to the Geometry selection in the
Details of “Fixed Support.” The top face of the L-shape is now fixed as shown below.
Step 8: Apply Loads
Right-click on Shape Optimization (A5). Choose Insert and then Force. In the Details of
“Force,” apply a 300 N force to the top edge of the free end in the Y-direction, as
shown below.
340
Finite Element Modeling and Simulation with ANSYS Workbench
Step 9: Define Target Weight Reduction
Click on Shape Finder underneath Solution (A6) in the Outline tree. In the Details of
“Shape Finder,” enter “60%” for the Target Reduction. Right-click on Solution (A6)
and select Solve.
Step 10: Retrieve Solution
Click on Shape Finder underneath Solution (A6) in the Outline tree to review results.
The region marked in gray is suggested as the optimal shape for the design under
the given loads. Save project and exit the ANSYS Workbench.
Solution steps for portion (B):
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Lshape_DOE.wbpj.”
Design Optimization
341
Step 2: Create a Static Structural Analysis System
Drag the Static Structural icon from the Analysis Systems Toolbox window and drop it
inside the highlighted green rectangle in the Project Schematic window.
Step 3: Launch the DesignModeler Program
Double-click the Geometry cell to launch DesignModeler, and select “Millimeter” in the
Units pop-up window.
Step 4: Create the Geometry
Click on the Sketching tab. Draw a sketch of the L-shape on the XY Plane, as shown
below. An entity named Sketch1 will be shown underneath XY Plane of the model’s
Tree Outline.
342
Finite Element Modeling and Simulation with ANSYS Workbench
Apply fillets of radius 1 mm to the six interior corners as shown below. Apply the Equal
Radius constraint to all six fillets. Turn on the Name option for the Display of Dimensions.
In the Details View, click on the box to the left of the dimension name V5. In the
pop-up window, enter “height1” as the parameter name and then click OK. Repeat
the steps for the dimension name V15 and enter “height2” as the corresponding
parameter name (note your dimension names may not be the same as they appear
here). A letter D will be marked on the box to the left of the two selected dimensions in the Details View as shown below.
Design Optimization
Extrude Sketch1 to create a 2 mm thick solid body, and close DesignModeler.
Note that a Parameter Set bus is added to the project schematic.
343
344
Finite Element Modeling and Simulation with ANSYS Workbench
Double-click on the Parameter Set. Note that the selected dimension parameters,
namely, height1 and height2, are now included as two design variables in the Outline
window.
Step 5: Launch the Static Structural Program
Double-click on the Model cell to launch the Static Structural program. Change the
Units to Metric (mm, kg, N, s, mV, mA).
Design Optimization
345
Step 6: Generate Mesh
Click on Mesh in the Outline tree. In the Details of “Mesh,” enter “1.0 mm” for the
Element Size. Right-click on Mesh and select Generate Mesh.
Step 7: Apply Boundary Conditions
Right-click on Static Structural (A5). Choose Insert and then Fixed Support from the
context menu. Click on the top face, and apply it to the Geometry selection in the
Details of “Fixed Support.” The top face of the L-shape is now fixed as shown below.
346
Finite Element Modeling and Simulation with ANSYS Workbench
Step 8: Apply Loads
Right-click on Static Structural (A5). Choose Insert and then Force. In the Details of
“Force,” apply a 300 N force to the bottom edge on the right along negative
Y-direction, as shown below.
Step 9: Retrieve Solution and Define Response Parameters
Insert Total Deformation by right-clicking Solution (A6) in the Outline tree. Right-click
on Solution (A6) and select Solve. Click on Total Deformation in the Outline to review
deformation results. Next, click on the box to the left of Maximum in the Details of
“Total Deformation.” Note that a letter “P” is added to the box.
Design Optimization
347
In the Details of “Solid,” click to the left of Mass under Properties. Note that a letter “P”
is added to the box. Exit the Static Structural window.
Step 10: Explore the Parameter Set
Double-click the Parameter Set in the project schematic. The input parameters (design
variables) and output parameters (responses) are listed with their corresponding
base values in the Outline window. In the Table of Design Points, the current design
is shown as the only design point, click Return to Project.
348
Finite Element Modeling and Simulation with ANSYS Workbench
Step 11: Create Response Surface Optimization
Double-click the Response Surface Optimization underneath Design Exploration from
the Toolbox. A Response Surface Optimization system will be added to the project
schematic. The optimization system is linked to the Parameter Set indicating that
it has access to the design and response parameters. Double-click on the Design of
Experiments tab in the project schematic.
Step 12: Generate DOE Table
In the Outline of Schematic B2: Design of Experiments, click on the P1-height1 tab. Enter
the Lower Bound and Upper Bound values as shown below.
Design Optimization
349
In the Outline of Schematic B2: Design of Experiments, click on the P2-height2 tab. Enter
the Lower Bound and Upper Bound values as shown below in the Properties of
Schematic B2: Design of Experiments.
In the Outline of Schematic B2: Design of Experiments, click on the Design of Experiments
tab. Turn on Preserve Design Points After DX Run. Change the Design Type to
Face-Centered.
350
Finite Element Modeling and Simulation with ANSYS Workbench
In the toolbar, click Preview. A DOE table listing nine design points is generated.
Then click Update in the toolbar to run finite element simulations for the nine different design scenarios. It will take a while to finish running all the simulations.
After completion, review the nine design points (simulation results) listed in the
Table of Schematic B2, and click Return to Project.
Step 13: Build Response Surface
Double-click the Response Surface tab in the project schematic. Turn on the two options
underneath Design Points in the Properties of Schematic B3: Response Surface. Click
Update to create the response surface.
Design Optimization
351
To view the response surface, click on the Response tab in the Outline of Schematic
B3: Response Surface. Select 3D for Mode in the Properties of Outline: Response. The
response surface of the maximum total deformation as a function of height1 and
height2 is shown in the Response Chart. Review the response surface chart, and
click Return to Project.
Step 14: Set Up Objectives and Constraints
Double-click the Optimization tab in the project schematic. Click on the Objectives and
Constraints tab in the Outline of Schematic B4: Optimization. Set up the objective and
constraints as shown below in the Table of Schematic B4: Optimization.
352
Finite Element Modeling and Simulation with ANSYS Workbench
Click on the Optimization tab in the Outline of Schematic B4: Optimization. Select
Screening for Optimization Method in the Chart: Response Chart for P3. Click
Update.
Three candidates for best design are now listed in the Table of Schematic B4:
Optimization. Review the best design candidates in the Candidate Points table. Click
Return to Project.
Design Optimization
353
Step 15: Update Current Design
Double-click the Parameter Set in the project schematic. In the Table of Design Points
shown below, DP1, DP2, and DP3 are feasible solutions that meet the design
requirements. Among them, DP3 achieves the best result in minimizing the total
weight while meeting the maximum total deformation requirement. It has a value
of 21 mm for height1 and 8 mm for height2.
We will now update the current design with this design point (DP3), which can then
be served as the base case for another round of optimization iteration if necessary.
Right-click on the DP3 tab in the Table of Design Points, and select Copy inputs to
Current in the drop-down menu. This will update the Current design point with
the selected design point DP3.
354
Finite Element Modeling and Simulation with ANSYS Workbench
Right-click on the Current tab in the Table of Design Points, and select Update Selected
Design Points. This will update the results of the Current design point. Click Return
to Project.
Double-click the Geometry tab in the project schematic to open DesignModeler. Note
that the geometry displayed in the Graphics window is now updated based
on the current design point (21 mm for height1 and 8 mm for height2). Close the
DesignModeler.
Design Optimization
355
Double-click the Results tab in the project schematic to view the updated deformation
results from the Static Structural–Mechanical program.
11.6 Summary
In this chapter, we briefly discussed design optimization problems and the type of
optimizations. We also learnt how to apply FEA in design optimization studies. A case
study is demonstrated using ANSYS Workbench to reduce the weight of an L-shaped
structure.
11.7 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1. Understand the basic concepts in design optimization and the steps in conducting
an optimization of a design with the FEA
2. Perform optimization of structure designs using ANSYS Workbench
356
Finite Element Modeling and Simulation with ANSYS Workbench
PROBLEMS
11.1 The 3-D design space of a steel bridge structure is 6 m long, 2 m tall, and 1 m wide,
as shown below. Suppose the structure is fixed on the bottom two edges and is
applied a uniform pressure load on the top surface. Perform topology optmization of the bridge using the given design space to achieve 80% weight reduction.
11.2 The following 1 m × 1 m square plate with a center hole of diameter 0.4 m is made
of structural steel and has a thickness of 0.1 m. Suppose the plate is fixed on the
left upper and lower edges, and is applied a downward force of 1000 N on the
upper right edge. Perform topology optmization of the plate to achieve a 60% and
90% weight reduction, respectively.
11.3 Optimize the stiffened aluminum panel design with clamped edges as shown
below. In the initial design, the 100 in. × 100 in. panel has a thickness of 0.5 in. The
stiffeners are each 100 in. long, 5 in. tall, and 0.25 in. thick, and divide the panel
evenly into eight blocks of the same area. The design variables, that is, the panel
thickness t, the longitudinal stiffener thickness tlong, and the lateral stiffener thickness tlat, have the following range of variations: 0.4 in. ≤ t ≤ 0.6 in., 0.2 in. ≤ tlong
≤ 0.3 in., and 0.2 in. ≤ tlat ≤ 0.3 in. The optimization aims to minimize the
panel’s weight subject to the constraint of the panel’s fundamental frequency
f base ≥ 130 Hz.
Panel thickness t
Longitudinal stiffener
thickness tlong
Lateral stiffener
thickness tlat
357
Design Optimization
11.4 Optimize the design of the following L-shaped bracket made of structural
steel with Young’s modulus E = 200 GPa and Poisson’s ratio ν = 0.3. In the initial design, the L-shape has a length L = 8 cm, a height h = 3 cm, and an extrusion depth d = 4 cm, and is rigidly connected to a rectangular block of size
3 cm × 8 cm × 1 cm, as shown below. The bracket is fixed on the right end. A
uniform pressure of 100 N/cm2 is applied on the protruding top face on the left.
The design variables, that is, L, h, and d, have the following range of variations:
6 cm ≤ L ≤ 8 cm, 2.5 cm ≤ h ≤ 3 cm, and 3 cm ≤ d ≤ 4 cm. The optimization aims to
minimize the bracket weight subject to the maximum allowed displacement limit
of 0.1 mm and stress limit of 100 MPa.
All dimensions are in
centimeters.
1.000
Depth d
R1.000
Height h
1.000
3.000
8.000
Length L
11.5 Perform topology and size optimization of the following bookshelf bracket
design. The bracket is 2 mm thick and is made of structural steel. The vertical
side of the bracket is fixed, and the horizontal side is applied a uniform pressure
of 25 MPa. (a) The objective of the topology optimization is to achieve a weight
reduction of 60%. (b) Redesign the bracket based on the results from topology
optimization. For the parametric optimization of the redesigned bracket, minimize the bracket weight subject to the maximum allowed deformation constraint
of 0.5 mm and stress constraint of 100 MPa.
20 cm
20 cm
358
Finite Element Modeling and Simulation with ANSYS Workbench
11.6 The original design space along with the constraints and loading conditions is
given below for structural optimization of a bike frame. The coordinate dimensions are all in millimeters. Assume a 5 mm thickness for the design space and
perform topology optimization for the frame structure to achieve a weight reduction target of 60% and 80%, respectively. Use structural steel for the frame and
discuss on the results of optimization.
406 N
71 N
A (40, 83)
69.2 N
E (620, 83)
274 N
B (0, 0)
1960 N
D (874, –276)
C (439, 405)
200 N
The constraints and loads at the head,
seat, and pedal positions.
Initial design range coordinates
11.7 The following heat sink consists of a 1 m tall central tube and 12 radial fins
that are uniformly distributed every 30°C around the central tube. The central
tube has an outer diameter of 0.4 m and an inner diameter of 0.36 m. Each fin is
0.048 m wide, 0.4 m long, and 1 m tall. The vapor temperature inside the tube
is 315°C with a heat transfer coefficient of 2.94 × 105 W/(m2°C). The surrounding
air temperature is 0°C with a heat transfer coefficient of 40.88 W/(m2°C). The
heat sink is made of aluminum alloy with a thermal conductivity of 175 W/
(m K), a density of 2770 kg/m3, and a specific heat of 870 J/(kg K). Optimize
the fin width W and the fin length L to minimize the use of material while
achieving a minimum heat dissipation of 75 kW per unit length. The design
variables W and L have the following range of variations: 0.03 m < W < 0.052 m,
and 0.2 m < L < 0.45 m.
W
L
359
Design Optimization
11.8 The original design space along with the constraints and loading conditions is
given below for structural optimization of a chair. Assume that the chair is made
of Douglas fir with E = 104 MPa and ν = 0.3. The chair is subjected to two surface
loads with force magnitudes of 980 and 98 N, respectively, as shown in the figure.
The four corners at the bottom of the chair are fixed. Optimize the chair design to
achieve a 90% weight reduction.
Surface B
FB= 98 N
1.4 m
FA= 980 N
0.6 m
Surface A
0.6 m
0.8 m
0.6 m
12
Failure Analysis
12.1 Introduction
Structures such as bridges, aircraft, and machine components can fail in many different
ways. An overloaded structure may experience permanent deformation, which can lead
to compromised function or failure of the entire structure. When subjected to millions of
small repeated loads, a structure may have a slow growth of surface cracks that can cause
material strength degradation and a sudden failure. When a slender structure is loaded
in compression, it may undergo an unexpected large deformation and lose its ability to
carry loads. Failure analysis plays an important role in improving the safety and reliability of an engineering structure. In this chapter, we will discuss some of the topics related
to structural failure analysis. The concepts of static, fatigue, and buckling failures will be
introduced, along with examples using ANSYS® Workbench.
12.2 Static Failure
Under static loading conditions, material failure can occur when a structure is stressed
beyond the elastic limit. There are two types of material failures from static loading,
namely, ductile failure and brittle failure. The main difference between the two failure
types is the amount of plastic deformation a material experiences before fracture. As
illustrated in Figure 12.1, ductile materials tend to have extensive plastic deformation
before fracture, while brittle materials are likely to experience no apparent plastic deformation before fracture. For more information on ductile and brittle failure, please see
Reference [18].
12.2.1 Ductile Failure
In this section, two common theories on ductile failure, that is, the maximum shear stress
theory (or Tresca Criterion) and the distortion energy theory (or von Mises Criterion), are
reviewed.
12.2.1.1 Maximum Shear Stress Theory (Tresca Criterion)
According to the maximum shear stress theory, ductile failure occurs when the maximum
shear stress τ max exceeds one half of the material yield strength Sy. Suppose that a factor of
safety n is considered in the design. The design equation is given as
τ max <
Sy
2n
(12.1)
361
362
Finite Element Modeling and Simulation with ANSYS Workbench
Brittle failure
Stress (σ)
Ductile failure
Ductile
fracture after
necking
Strain (ε)
Brittle
fracture
FIGURE 12.1
Stress–strain curves for brittle and ductile materials.
12.2.1.2 Distortion Energy Theory (von Mises Criterion)
Based on the distortion energy theory, ductile failure occurs when the maximum von
Mises stress σ e max exceeds the material yield strength Sy. Applying a factor of safety n to
the design, the design equation becomes:
σ e max <
Sy
n
(12.2)
12.2.2 Brittle Failure
For brittle materials, the two common failure theories are the maximum normal stress
theory and the Mohr–Coulomb theory.
12.2.2.1 Maximum Normal Stress Theory
The maximum normal stress theory states that brittle failure occurs when the maximum
principal stress exceeds the ultimate tensile (or compressive) strength of the material.
Suppose that a factor of safety n is considered in the design. The safe design conditions
require that
−
Suc
S
< {σ 1 , σ 2 , σ 3 } < ut
n
n
(12.3)
where σ1, σ2, σ3 are the principal stresses in the three directions, Sut and Suc are the ultimate
tensile strength and the ultimate compressive strength, respectively. To avoid brittle failure, the principal stresses at any point in a structure should lie within the square failure
envelope illustrated in Figure 12.2 based on the maximum normal stress theory.
12.2.2.2 Mohr–Coulomb Theory
The Mohr–Coulomb theory predicts brittle failure by comparing the maximum principal
stress with the ultimate tensile strength Sut and the minimum principal stress with the
363
Failure Analysis
σ3
Sut
n
σ1
Suc
n
Sut
n
Suc
n
FIGURE 12.2
The square failure envelope for the maximum normal stress theory.
σ3
Sut
n
Sut
n
Suc
n
σ1
Suc
n
FIGURE 12.3
The hexagonal failure envelope for the Mohr–Coulomb theory.
ultimate compressive strength Suc. Suppose that a factor of safety n is considered in the
design. The safe design conditions require that the principal stresses lie within the hexagonal failure envelope illustrated above in Figure 12.3.
12.3 Fatigue Failure
Under dynamic or cyclic loading conditions, fatigue can occur at stress levels that are
considerably lower than the yield or ultimate strengths of the material. Fatigue failures are
364
Finite Element Modeling and Simulation with ANSYS Workbench
Stress
σa
σmax
σmin
σm
Cycles
FIGURE 12.4
A typical fatigue stress cycle.
often sudden with no advanced warning and can lead to catastrophic results. In this section, theories related to fatigue failure are reviewed. For more details, we refer the reader
to Reference [18].
Design of components subjected to cyclic load involves the concept of mean and alternating stresses (see Figure 12.4 for an illustration of a typical fatigue stress cycle). The
mean stress, σm, is the average of the maximum and minimum stresses in one cycle, that
is, σ m = (σ max + σ min )/2. The alternating stress, σa, is one-half of the stress range in one
cycle, that is, σ a = (σ max − σ min )/2. For most materials, there exist a fatigue limit, and
parts having stress levels below this limit are considered to have infinite fatigue life. The
fatigue limit is also referred to as the endurance limit, Se.
In the following, three commonly used fatigue failure theories, namely, Soderberg,
Goodman, and Gerber failure criteria, are presented.
12.3.1 Soderberg Failure Criterion
This theory states that the structure is safe if
σa σm
1
+
<
Se
Sy
n
(12.4)
where σa is the alternating stress, σm the mean stress, Se the endurance limit, Sy the yield
strength, and n the factor of safety. When the alternating stress σa is plotted versus the
mean stress σm, the Soderberg line can be drawn between the points of σ a = Se /n and
σ m = Sy/n . If the stress is below the line, then the design is safe. This is a conservative
criteria based on the material yield strength Sy (see Figure 12.5).
12.3.2 Goodman Failure Criterion
This theory states that the structure is safe if
σa σm
1
+
<
Se
Sut
n
(12.5)
where Sut is the ultimate tensile strength and n is the factor of safety. When the alternating
stress σa is plotted versus the mean stress σm, the Goodman line can be drawn between
365
Failure Analysis
σa
Soderberg line
Se
n
Goodman line
Gerber parabola
Sy
n
Sut
n
σm
FIGURE 12.5
Fatigue diagrams showing various criteria of failure.
the points of σ a = Se/n and σ m = Sut/n. If the stress is below the line, then the design is
safe. This is a less conservative criteria based on the material ultimate strength Sut (see
Figure 12.5).
12.3.3 Gerber Failure Criterion
This theory states that the structure is safe if
2
σa σm
1
+
<
Se Sut
n
(12.6)
where Sut is the ultimate tensile strength and n is the factor of safety. When the alternating
stress σa is plotted versus the mean stress σm, the Gerber parabola can be drawn between the
points of σ a = Se /n and σ m = Sut /n. If the stress is below the parabola, then the design is safe.
This is a less conservative criteria based on the material ultimate strength Sut (see Figure 12.5).
In general, most structures will fail eventually after a certain number of cycles of
repeated loading. In fatigue life analysis, S–N curves are widely used for the life estimation, especially in high-cycle fatigue situations where the cyclic loading is elastic and the
number of cycles to failure is large. An S–N curve (see Figure 12.6) relates the alternating
stress, σa, to the number of cycles to failure, Nf, and is generated from fatigue tests on
samples of the given material.
Most S–N curves are developed based on zero mean stress. However, many loading
applications involve a nonzero mean stress. The Soderberg, Goodman, and Gerber theories mentioned above can be used to estimate the mean stress effects on fatigue life. Note
that there is little difference in the three theories when the mean stress σm is relatively
small compared to the alternating stress σa.
Take the Gerber theory as an example. To estimate a structure’s fatigue life for a given
combination of mean stress σm and alternating stress σa, one may first solve for the effective
alternating stress σe, based on the following modified Gerber equation:
2
σa σm
1
+
=
σ e Sut
n
(12.7)
Finite Element Modeling and Simulation with ANSYS Workbench
Stress (σa)
366
σe
Endurance limit
Se
Fatigue life at
σe
Cycles to
failure (Nf)
FIGURE 12.6
Use of the S–N curve for fatigue life estimation.
Once the effective alternating stress, σe, is obtained from Equation 12.7, the fatigue life
for the corresponding effective alternating stress σe can be determined from the S–N curve
of the given material (see Figure 12.6). If the effective alternating stress, σe, is kept below
threshold of the endurance limit, Se, the structure is considered to have an infinite life.
12.4 Buckling Failure
Under compressive loads, slender members such as columns can become structurally
unstable, leading to a sudden failure. The sudden loss of stability under compressive loads
is referred to as buckling [19]. Buckling failure usually occurs within the elastic range of
the material. The actual compressive stress in the structure at failure is often smaller than
the ultimate compressive strength of the material.
In buckling analysis, one may be interested in finding the critical load, that is, the
maximum compressive load a structure can support before buckling occurs, or the
buckling mode shapes, that is, shapes the structure takes while buckling. The critical load is a structure’s elastic stability limit, which is adequate to force the structure
to buckle into the first mode shape. A structure can potentially have infinitely many
buckling modes, and higher critical load values are required to trigger higher-order
buckling modes (see Figure 12.7 for an illustration of buckling failure of a slender column, and Figure 12.8 for different buckling mode shapes). Because buckling failure can
lead to catastrophic results, a high factor of safety (at least >3) is often used for critical
load calculations.
When developing a model for buckling, one should pay attention to the boundary conditions as they have a considerable effect on a structure’s buckling behavior. For example,
a column with both ends pinned can buckle more freely than a column with both ends
fixed. Also note that bending moments acting on a structural member, for example, due
to e­ ccentricity of an axial load, may play a role in affecting buckling and should not be
neglected in the analysis in general.
367
Failure Analysis
(a)
(b)
FIGURE 12.7
Buckling of a slender column: (a) load < critical load, (b) load > critical load. (http://en.wikipedia.org/wiki/
Buckling.)
(a)
(b)
(c)
FIGURE 12.8
Buckling of a slender column: (a) unbuckled, (b) first buckling mode, and (c) second buckling mode.
12.5 Case Studies with ANSYS Workbench
Problem Description: A dog-bone shaped specimen is examined for static, fatigue, and
buckling failures. The specimen is made of structural steel with geometric dimensions
shown below. The bottom face of the specimen is fixed, and the top face of the specimen
is applied a static pressure load of 50 MPa. (a) Determine whether or not the specimen
undergoes plastic deformation under the given static pressure load. (b) If the static pressure load is changed into a fully reversed cyclic load with a magnitude of 50 MPa, find
the life of the specimen, and also determine whether or not fatigue failure occurs in the
specimen assuming a design life of 106 cycles. (c) Determine whether or not the specimen
buckles under the given static pressure load, and obtain the first three buckling mode
shapes.
368
Finite Element Modeling and Simulation with ANSYS Workbench
15.000
R3.000
38.000
The specimen
thickness is
0.75 mm.
8.000
5.000
All dimensions are in mm.
Solution steps for portion (A and B):
Step 1: Start an ANSYS Workbench Project
Launch ANSYS Workbench and save the blank project as “Dogbone.wbpj.”
Step 2: Create a Static Structural Analysis System
Drag the Static Structural icon from the Analysis Systems Toolbox window and drop it
inside the highlighted green rectangle in the Project Schematic window.
Step 3: Launch the DesignModeler Program
Double-click the Geometry cell to launch DesignModeler, and select “Millimeter” in the
Units pop-up window.
Failure Analysis
369
Step 4: Create the Geometry
Click on the Sketching tab. Draw a sketch of the dog bone shape on the XY Plane, as
shown below. An entity named Sketch1 will be shown underneath XY Plane of the
model’s Tree Outline.
Extrude Sketch1 to create a 0.75 mm thick solid body, as shown below.
370
Finite Element Modeling and Simulation with ANSYS Workbench
Step 5: Launch the Static Structural Program
Double-click on the Model cell to launch the Static Structural program. Change the
Units to Metric (mm, kg, N, s, mV, mA).
Step 6: Generate Mesh
Click on Mesh in the Outline tree. In the Details of “Mesh,” enter “0.5 mm” for the
Element Size. Right-click on Mesh and select Generate Mesh.
Failure Analysis
371
Step 7: Apply Boundary Conditions
Right-click on Static Structural (A5). Choose Insert and then Fixed Support from the
context menu. Click on the bottom face, and apply it to the Geometry selection in
the Details of “Fixed Support.” The bottom face of the dog bone shape is now fixed
as shown below.
Step 8: Apply Loads
Right-click on Static Structural (A5). Choose Insert and then Pressure. In the Details of
“Pressure,” apply a 50 MPa pressure to the top face, as shown below.
372
Finite Element Modeling and Simulation with ANSYS Workbench
Step 9: Retrieve Static Analysis Results
First, insert a Total Deformation item by right-clicking on Solution (A6) in the project
Outline.
Then, insert an Equivalent Stress item by right-clicking on Solution (A6) in the project
Outline.
Next, right-click on Solution (A6) in the project Outline, and select Insert -> Stress Tool
-> Max Equivalent Stress. The initial yielding in the test sample may be predicted by
comparing the maximum von Mises stress in the specimen with the tensile yield
strength of the specimen material. The Stress Tool is used here to show the safety
factor results.
Right-click on Solution (A6) and select Solve. The computed total deformation, von
Mises stress and safety factor distributions are shown below. From the static analysis results, it is apparent that the neck portion of the specimen will not yield if
loaded statically.
A: Static structural
Total deformation
Type: Total deformation
Unit: mm
Time: 1
0.022731 Max
0.020205
0.01768
0.015154
0.012628
0.010103
0.007577
0.0050513
0.0025257
0 Min
A: Static structural
Equivalent stress
Type: Equivalent (von Mises) stress
Unit: Mpa
Time: 1
216.97 Max
192.87
168.77
144.68
120.58
96.488
72.392
48.297
24.201
0.10569 Min
A: Static structural
Safety factor
Type: Safety factor
Time: 1
15 Max
10
5
1.1523 Min
0
Step 10: Retrieve Fatigue Analysis Solution
Right-click on Solution (A6) in the Outline, and select Insert -> Fatigue -> Fatigue Tool.
In the Details of “Fatigue Tool,” set the Mean Stress Theory to Goodman. Note that the
default loading type is Fully Reversed constant amplitude load, and that the default
analysis type is the Stress Life type using the von Mises stress calculations.
373
Failure Analysis
Right-click on the Fatigue Tool in the Outline, and select Insert -> Life. Next, right-click
on the Fatigue Tool and select Insert -> Safety Factor. In the Details of “Safety Factor,”
change the Design life from the default value of 109 cycles to 106 cycles. Finally, rightclick on the Fatigue Tool and select Evaluate All Results. From the fatigue analysis
results, the shortest life is at the undercut fillets (19,079 cycles) followed by the neck
portion of the specimen. The neck portion of the specimen has a fairly small safety
factor with a minimum value of 0.3973. The results show that the specimen will not
survive the fatigue testing assuming a design life of 106 cycles.
A: Static structural
Life
Type: Life
Time: 0
1e6 Max
6.441e5
4.1486e5
2.6721e5
1.7211e5
1.1086e5
71,402
45,990
29,622
19,079 Min
A: Static structural
Safety factor
Type: Safety factor
Time: 0
15 Max
10
5
0.3973 Min
0
374
Finite Element Modeling and Simulation with ANSYS Workbench
Solution steps for portion (C):
Step 1: Create a Linear Buckling Analysis System
In the Project Schematic window, right-click on the Solution cell of the Static Structural
analysis system and select Transfer Data to New -> Linear Buckling. A linear buckling analysis system will be added, with the static structural results being used as
initial conditions. The engineering data, geometry, and model will be shared by
both analyses.
Step 2: Launch the Multiple Systems–Mechanical Program
Double-click the Setup cell of the Linear Buckling system to launch the Multiple Systems–
Mechanical program. Click on the Analysis Settings under Linear Buckling (B5) in the
Outline. In the Details of “Analysis Settings,” set the Max Modes to Find to 3.
Failure Analysis
375
Step 3: Retrieve Linear Buckling Analysis Results
Insert three Total Deformation items by right-clicking on Solution (B6) in the Outline. In
the Details of “Total Deformation,” set Mode to 1. In the Details of “Total Deformation
2,” set Mode to 2. In the Details of “Total Deformation 3,” set Mode to 3.
Right-click on Solution (B6) and select Solve to view the buckling modes. To use the
default window layout as shown below, select View -> Windows -> Reset Layout
from the top menu bar. Note that the load modifier for the first buckling mode is
found to be 0.78173. To find the load required to buckle the structure, multiply the
applied load by the load multiplier. For example, the first buckling load will be
39.0865 MPa (0.78173 × 50 MPa), thus the applied pressure of 50 MPa will cause the
specimen to buckle. In the Graph window, you can play the buckling animation.
376
Finite Element Modeling and Simulation with ANSYS Workbench
The following figures show the first three buckling mode shapes. The corresponding
load multipliers for the first, second, and third mode shapes are 0.78173, 2.0094,
and 5.7321, respectively.
B: Linear buckling
Total deformation
Type: Total deformation
Load multiplier: 0.78173
Unit: mm
1 Max
0.88889
0.77778
0.66667
0.55556
0.44444
0.33333
0.22222
0.11111
0 Min
B: Linear buckling
Total deformation 2
Type: Total deformation
Load multiplier: 2.0094
Unit: mm
1 Max
0.88889
0.77778
0.66667
0.55556
0.44444
0.33333
0.22222
0.11111
0 Min
B: Linear buckling
Total deformation 3
Type: Total deformation
Load multiplier: 5.7321
Unit: mm
1 Max
0.88889
0.77778
0.66667
0.55556
0.44444
0.33333
0.22222
0.11111
0 Min
Note that the max value in the total deformation plots is scaled to 1 when displaying the buckling mode shapes. Here, the deformation plot is used for mode shape
visualization, with the actual values of deformation carrying no physical meaning.
12.6 Summary
In this chapter, we briefly reviewed the static material failure theories (the ductile failure
and brittle failure theories), the fatigue failure theories for structures under low and cyclic
loading, and the buckling failure theory. An example of a dog-bone shaped specimen is
used to demonstrate how to conduct the static, fatigue, and buckling failure analyses using
the FEM with ANSYS Workbench.
377
Failure Analysis
12.7 Review of Learning Objectives
Now that you have finished this chapter you should be able to
1. Understand the failure theories for structures under static load
2. Understand the fatigue failure theories for structures under cyclic load
3. Understand the buckling failure of a slender structure under static compressive
load
4. Perform static failure analysis of structures using ANSYS Workbench
5. Perform fatigue failure analysis of structures using ANSYS Workbench
6. Perform buckling failure analysis of structures using ANSYS Workbench
PROBLEMS
12.1 The following figure shows a simple bracket made of structural steel with Young’s
modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The bracket has a front face profile shown below and an extrusion depth of 12 cm. A through-hole of diameter
1.5 cm is centered on the bracket’s flat ends, one on each side. Suppose that the
bracket is fixed on the bottom, and is subjected to a fully reversed cyclic pressure
load of magnitude 10 MPa on the top face. Find the life of the bracket and determine whether or not fatigue failure occurs in the part assuming a design life of
106 cycles.
Cyclic loading
All dimensions are in
centimeters.
10.000
1.500
1.500
R3.000
1.500
8.000
1.500
12.000
8.000
1.500
378
Finite Element Modeling and Simulation with ANSYS Workbench
12.2 The following figure shows a lifting bracket made of structural steel with Young’s
modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The base portion of the bracket
is 20 cm long, 12 cm wide, and 3 cm deep. The fillet radius at the base corners is
1.75 cm. Four through-holes each of radius 1 cm are located symmetrically both
length wise and width wise in the base, with the hole centers being 12 cm apart by
6 cm apart. The upper portion of the base has a profile sketch shown below and is
3 cm thick. The bracket is fixed on the bottom. (1) Suppose the bracket is lifted by
a bearing load of 100 kN. Determine if the bracket can safely carry the given load.
(2) Find the life of the bracket if it is subjected to a fully reversed cyclic bearing
load of magnitude 50 kN.
R3.000
R1.500
All dimensions are in
centimeters.
3.000
3.000
8.000
6.000
6.000
12.3 The figure below shows a bracket made of structural steel with Young’s modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The rectangular base of the bracket is
12 cm long, 6 cm wide, and 3 cm deep. The midsection has an overall extrusion
depth of 1.6 cm and a profile sketch shown below. The hollow cylinder has an
inner radius of 1 cm, an outer radius of 2 cm, and a height of 4 cm. The central axis
of the cylinder is 12.5 cm away from the leftmost end of the bracket. The bracket
is fixed on the left end. (1) Determine if the bracket can safely carry the load when
a pressure of 800 N/cm2 is exerted on the inclined face of the midsection. (2) Find
the life of the bracket if it is subjected to a fully reversed cyclic pressure of magnitude 5 MPa on one side of the midsection as shown below.
All dimensions are in
centimeters.
6.000
Cyclic loading
8.000
4.000
379
Failure Analysis
12.4 The following figure shows an L-shaped bracket made of structural steel with
Young’s modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The 4 cm thick L-shape
with the cross-sectional profile shown in the figure is rigidly connected to a rectangular base of size 3 cm × 8 cm × 1 cm. The bracket is fixed on the right end at
the base. A surface pressure of 10 MPa is exerted on the left end as shown below.
(1) Determine if static failure or buckling failure occurs in the bracket. (2) Find the
life of the bracket if it is subjected to a fully reversed cyclic pressure of magnitude
10 MPa on the left end.
All dimensions are in
centimeters.
1.000
R1.000
Pressure loading
3.000
1.000
8.000
12.5 The following figure shows the cross-sectional profile of a 39 ft long straight railroad track. Use structural steel for the track with Young’s modulus E = 200 GPa
and Poisson’s ratio ν = 0.3. Assume the track is fixed on one end and subjected to a
pressure load along the axis of the track on the other end, that is, the cross section
at a distance of 39 ft from the fixed end. Determine the critical buckling load for
the railroad track.
1.375
R0.500
1.125
R0.500
105.00°
6.000
0.375
All length dimensions are
in inches.
R0.500
2.250
105.00°
0.375
380
Finite Element Modeling and Simulation with ANSYS Workbench
12.6 The following two-bar assembly consists of four components made of structural steel with Young’s modulus E = 200 GPa, Poisson’s ratio ν = 0.3, and density
ρ = 7850 kg/m3. As shown in the figure below, the two bars are identical in size.
Each bar has a diameter of 2 cm and a length of 16 cm. The two identical end
blocks each has a thickness of 2 cm and a fillet radius of 1 cm at the round corners. The distance of the gap in between the two blocks is 8 cm. The two bars are
fixed on the right ends, and subjected to a pressure load of 100 MPa on the left
ends as shown below. (1) Determine if static failure or buckling failure occurs in
the assembly. (2) Assuming a design life of 106, determine if fatigue failure occurs
in the assembly when subjected to a fully reversed cyclic pressure of magnitude
100 MPa on the left ends of the two bars.
Fixed ends
2.000 5.000
R1.000
Loading
10.000
6.000
All dimensions are in centimeters.
Use bonded for all contact regions.
12.7 The following figure shows a crank shaft made of structural steel with Young’s
modulus E = 200 GPa and Poisson’s ratio ν = 0.3. The crank shaft consists of three
identical horizontal rods and two identical vertical pieces. The rod is each 4 in. long.
The plate each has a thickness of 0.5 in. Along the four edges where the rods are connected with the plates, a fillet of radius 0.2 in. is added to create round corners. The
crank shaft is fixed on the two ends and subjected to a fully reversed cyclic pressure
of magnitude 300 psi on the top face of the central rod as shown below. Assuming a
design life of 106, determine if fatigue failure occurs in the crank shaft assembly.
All dimensions are in inches.
Cyclic loading
R1.250
Fixed end
D1.800
2.750
2.750
D1.800
R0.800
Fixed end
4.250
1.250
Appendix 1: Review of Matrix Algebra
A.1 Linear System of Algebraic Equations
a11x1 + a12 x2 + … + a1n xn = b1
a21x1 + a22 x2 + … + a2 n xn = b2
.......
an1x1 + an 2 x2 + … + ann xn = bn
(A.1)
where x1, x2, …, xn are the unknowns.
In matrix form:
Ax = b
(A.2)
where
a11
a
21
A = [aij ] =
…
an 1
a12
a22
…
an 2
…
…
…
…
a1n
a2 n
,
…
ann
x1
x
2
x = { xi } = ,
:
xn
b1
b
2
b = {bi } =
:
bn
(A.3)
A is called an n × n (square) matrix, and x and b are (column) vectors with dimension n.
A.2 Row and Column Vectors
v = [v1
v2
v3 ]
w1
w = w2
w
3
(A.4)
381
382
Appendix 1
A.3 Matrix Addition and Subtraction
For two matrices A and B, both of the same size (m × n), the addition and subtraction are
defined by
C = A + B with cij = aij + bij
D = A − B with dij = aij − bij
(A.5)
A.4 Scalar Multiplication
λA = [λ aij ]
(A.6)
A.5 Matrix Multiplication
For two matrices A (of size l × m) and B (of size m × n), the product of AB is defined by
m
C = AB with cij =
∑a b
ik kj
(A.7)
k =1
where i = 1, 2,…, l; j = 1, 2,…, n.
Note that, in general, AB ≠ BA, but (AB)C = A(BC) (associative).
A.6 Transpose of a Matrix
If A = [aij], then the transpose of A is
AT = [a ji ]
(A.8)
(AB)T = BT AT
(A.9)
Notice that
383
Appendix 1
A.7 Symmetric Matrix
A square (n × n) matrix A is called symmetric, if
A = AT
or aij = a ji
(A.10)
A.8 Unit (Identity) Matrix
1
0
I=
…
0
0
1
…
…
…
0
…
…
0
0
…
1
(A.11)
Note that AI = A, Ix = x.
A.9 Determinant of a Matrix
The determinant of square matrix A is a scalar number denoted by det A or |A|. For 2 × 2
and 3 × 3 matrices, their determinants are given by
a
det
c
b
= ad − bc
d
(A.12)
and
a11
det a21
a31
a12
a22
a32
a13
a23 = a11a22 a33 + a12 a23 a31 + a21a32 a13
a33
(A.13)
− a13 a22 a31 − a12 a21a33 − a23 a32 a11
A.10 Singular Matrix
A square matrix A is singular if det A = 0, which indicates problems in the system (non­
unique solutions, degeneracy, etc.).
384
Appendix 1
A.11 Matrix Inversion
For a square and nonsingular matrix A (det A ≠ 0), its inverse A−1 is constructed in such a
way that
(A.14)
AA −1 = A −1A = I
The cofactor matrix C of matrix A is defined by
Cij = (−1)i + j Mij
(A.15)
where Mij is the determinant of the smaller matrix obtained by eliminating the ith row and
jth column of A.
Thus, the inverse of A can be determined by
A −1 =
1
CT
det A
(A.16)
We can show that (AB)−1 = B−1A−1.
EXAMPLE A.1
a
c
(A)
b
d
−1
=
d
1
( ad − bc) −c
−b
a
Checking:
a
c
(B)
1
−1
0
−1
b a
d c
−1
2
−1
−b a
a c
b
d
1
=
d ( ad − bc) −c
0
−1
2
−1
3
1
=
2
( 4 − 2 − 1)
1
b 1
=
d 0
T
3
1
1 = 2
1
1
2
2
1
0
1
2
2
1
1
1
1
Checking:
1
−1
0
−1
2
−1
0 3
−1 2
2 1
2
2
1
1 1
1 = 0
1 0
0
1
0
0
0
1
If det A = 0 (i.e., A is singular), A−1 does not exist.
The solution of the linear system of equations (Equation 1.1) can be expressed as
(assuming the coefficient matrix A is nonsingular)
x = A −1b
Thus, the main task in solving a linear system is to find the inverse of the coefficient
matrix.
385
Appendix 1
A.12 Solution Techniques for Linear Systems of Equations
• Gauss elimination methods
• Iterative methods
The two methods are briefly reviewed in Chapter 5.
A.13 Positive-Definite Matrix
A square (n × n) matrix A is said to be positive definite, if for all nonzero vector x of dimension n,
x T Ax > 0
Note that positive-definite matrices are nonsingular. We can show that all stiffness
matrices are positive definite and the above condition is a statement that the strain energy
in a structure should be positive if the structure is constrained and the stiffness matrix is
nonsingular.
A.14 Differentiation and Integration of a Matrix
Let A(t) = [aij(t)], then the differentiation is defined by
d
daij (t)
A(t) =
dt
dt
(A.17)
∫ A(t)dt = ∫ a (t)dt
(A.18)
and the integration by
ij
Appendix 2: Photo Credits
Chapter 1
Figure 1.1(a) A train engine built with LEGO®.
(http://lego.wikia.com/wiki/10020_Santa_Fe_Super_Chief)
Figure 1.3(a) Wind load simulation of an offshore platform.
(Courtesy ANSYS, Inc. http://www.ansys.com/Industries/Energy/Oil+&+Gas);
Figure 1.3(b) Modal response of a steel frame building with concrete slab floors.
(http://www.isvr.co.uk/modelling/)
Figure 1.3(c) Underhood flow and thermal management.
(Courtesy ANSYS, Inc. http://www.ansys.com/Industries/Automotive/Application+
Highlights/Underhood)
Figure 1.3(d) Electric field pattern of antenna mounted on helicopter.
(Courtesy ANSYS, Inc. http://www.ansys.com/Industries/Electronics+&+Semiconduc
tor/Defense+&+Aerospace+Electronics)
Chapter 2
Figure 2.1(a) Montreal Biosphere Museum.
(http://en.wikipedia.org/wiki/Montreal_Biosph%C3%A8re)
Figure 2.1(b) Betsy Ross Bridge.
(http://en.wikipedia.org/wiki/Betsy_Ross_Bridge)
Chapter 3
Figure 3.1(a) A car frame.
(http://www.carbasics-1950.com/)
Figure 3.1(b) An exercise machine.
(http://www.nibbledaily.com/body-by-jake-cardio-cruiser/)
387
388
Appendix 2
Chapter 5
Figure 5.1(a) Reflective symmetry.
(http://www.thinkingfountain.org/s/symmetry/butterflypattern.gif)
Figure 5.1(c) Rotational symmetry.
(http://csdt.rpi.edu/na/pnwb/symmetry2a.html)
Figure 5.1(d) Translational symmetry.
(http://library.thinkquest.org/16661/background/symmetry.1.html)
Chapter 6
Figure 6.1 An airplane made of numerous plate and shell structures.
(http://en.wikipedia.org/wiki/Boeing_787_Dreamliner)
Chapter 8
Figure 8.7 Two equations for determining the proportional damping coefficients.
(From R. D. Cook, Finite Element Modeling for Stress Analysis (John Wiley & Sons, Inc.,
Hoboken, NJ, 1995).)
Figure 8.13 Car crash analysis using the FEA (from LS-Dyna website).
(http://www.dynaexamples.com/)
Chapter 10
Figure 10.1(a) Airflow around a tractor.
(Courtesy ANSYS, Inc. http://www.ansys.com/Industries/Automotive/Application+
Highlights/Body)
Figure 10.1(b) Streamlines inside a combustion chamber.
(Courtesy ANSYS, Inc. http://www.edr.no/en/courses/ansys_cfd_advanced_­
modeling_reacting_flows_and_combustion_in_ansys_fluent)
Chapter 12
Figure 12.7 Buckling of a slender column.
(http://en.wikipedia.org/wiki/Buckling)
References
1. O. C. Zienkiewicz, The birth of the finite element method and of computational mechanics,
International Journal for Numerical Methods in Engineering, 60, 3–10, 2004.
2. ANSYS Workbench User’s Guide, Release 14.5, ANSYS, Inc., 2012.
3. C. Felippa, Introduction to Finite Element Methods, Lecture Notes, University of Colorado at
Boulder (http://www.colorado.edu/engineering/cas/courses.d/IFEM.d/).
4. J. N. Reddy, An Introduction to the Finite Element Method, 3rd ed. (McGraw-Hill, New York, 2006).
5. O. C. Zienkiewicz and R. L. Taylor, The Finite Element Method, 5th ed. (McGraw-Hill, New York,
2000).
6. S. P. Timoshenko and S. Woinowsky-Krieger, Theory of Plates and Shells, 2nd ed. (McGraw-Hill,
New York, 1987).
7. R. D. Cook, Finite Element Modeling for Stress Analysis (John Wiley & Sons, Inc., Hoboken, NJ,
1995).
8. K. J. Bathe, Finite Element Procedures (Prentice-Hall, Englewood Cliffs, NJ, 1996).
9. T. R. Chandrupatla and A. D. Belegundu, Introduction to Finite Elements in Engineering, 3rd ed.
(Prentice-Hall, Upper Saddle River, NJ, 2002).
10. Y. C. Fung, A First Course in Continuum Mechanics, 3rd ed. (Prentice-Hall, Englewood Cliffs, NJ,
1994).
11. P. Timoshenko and J. N. Goodier, Theory of Elasticity, 3rd ed. (McGraw-Hill, New York, 1987).
12. ANSYS Help System, Mechanical Applications—Mechanical User Guide, Release 14.5, ANSYS,
Inc., 2012.
13. R. D. Cook, D. S. Malkus, M. E. Plesha, and R. J. Witt, Concepts and Applications of Finite Element
Analysis, 4th ed. (John Wiley & Sons, Inc., Hoboken, NJ, 2002).
14. S. Moaveni, Finite Element Analysis—Theory and Application with ANSYS, 3rd ed. (Prentice-Hall,
Upper Saddle River, NJ, 2007).
15. B. R. Munson, D. F. Young, and T. H. Okiishi, Fundamentals of Fluid Mechanics, 5th ed. (John
Wiley & Sons, Inc., Hoboken, NJ, 2006).
16. ANSYS Help System, CFX—Modeling Guide, Release 14.5, ANSYS, Inc., 2012.
17. ANSYS Help System, Design Exploration User Guide, Release 14.5, ANSYS, Inc., 2012.
18. J. E. Shigley, C. R. Mischke, and R. G. Budynas, Mechanical Engineering Design, 7th ed. (McGrawHill, New York, 2003).
19. F. P. Beer, E. R. Johnston, J. T. DeWolf,x and D. F. Mazurek, Mechanics of Materials, 6th ed.
(McGraw-Hill, New York, 2011).
389
Mechanical Engineering
Finite Element Modeling and Simulation
with
“This book covers a wide range of engineering analyses using FEM with adequate
theoretical background, illustrative examples, and detailed tutorials on using ANSYS
Workbench for FEA. ... It covers both the theoretical and practical parts of FEM, which
is very helpful to engineering students.”
—Dr. Anh-Vu Phan, University of South Alabama, Mobile, USA
“The book is an excellent resource to introduce undergraduate and graduate engineering students to finite element analysis. A great deal of emphasis is [placed] on the
finite element methods through simple application examples. …The main thing that I
like about this book is that the authors have presented the material using ANSYS
Workbench. For undergraduate students who are considered beginners with FEA, this
is an ideal software to use. The examples are simple and clear, and the ANSYS Workbench case studies are good applications of the techniques.”
—Dr. Mohammad H. Alimi, California State University, Fresno, USA
“This book should be a great resource for engineers or senior undergraduate students
who want to learn FEA and use ANSYS themselves and therefore can be adopted as a
textbook for senior undergraduate students in engineering when introducing FEA.”
—Dr. Ernian Pan, University of Akron, Ohio, USA
Learn Basic Theory and Software Usage from a Single Volume
Finite Element Modeling and Simulation with ANSYS Workbench combines finite
element theory with real-world practice. Providing an introduction to finite element
modeling and analysis for those with no prior experience, and written by authors with
a combined experience of 30 years teaching the subject, this text presents FEM
formulations integrated with relevant hands-on applications using ANSYS Workbench
for finite element analysis (FEA).
Incorporating the basic theories of FEA and the use of ANSYS Workbench in the
modeling and simulation of engineering problems, the book also establishes the
FEM method as a powerful numerical tool in engineering design and analysis.
This text benefits upper-level undergraduate students in all engineering disciplines,
as well as researchers and practicing engineers who use the finite element method
to analyze structures.
K13420
ISBN: 978-1-4398-7384-7
90000
9 781439 873847
